1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Changing options at runtime with a command
318 Some options can be changed during the operation of the filter using
319 a command. These options are marked 'T' on the output of
320 @command{ffmpeg} @option{-h filter=<name of filter>}.
321 The name of the command is the name of the option and the argument is
325 @chapter Options for filters with several inputs (framesync)
326 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
328 Some filters with several inputs support a common set of options.
329 These options can only be set by name, not with the short notation.
333 The action to take when EOF is encountered on the secondary input; it accepts
334 one of the following values:
338 Repeat the last frame (the default).
342 Pass the main input through.
346 If set to 1, force the output to terminate when the shortest input
347 terminates. Default value is 0.
350 If set to 1, force the filter to extend the last frame of secondary streams
351 until the end of the primary stream. A value of 0 disables this behavior.
355 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
357 @chapter Audio Filters
358 @c man begin AUDIO FILTERS
360 When you configure your FFmpeg build, you can disable any of the
361 existing filters using @code{--disable-filters}.
362 The configure output will show the audio filters included in your
365 Below is a description of the currently available audio filters.
369 A compressor is mainly used to reduce the dynamic range of a signal.
370 Especially modern music is mostly compressed at a high ratio to
371 improve the overall loudness. It's done to get the highest attention
372 of a listener, "fatten" the sound and bring more "power" to the track.
373 If a signal is compressed too much it may sound dull or "dead"
374 afterwards or it may start to "pump" (which could be a powerful effect
375 but can also destroy a track completely).
376 The right compression is the key to reach a professional sound and is
377 the high art of mixing and mastering. Because of its complex settings
378 it may take a long time to get the right feeling for this kind of effect.
380 Compression is done by detecting the volume above a chosen level
381 @code{threshold} and dividing it by the factor set with @code{ratio}.
382 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
383 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
384 the signal would cause distortion of the waveform the reduction can be
385 levelled over the time. This is done by setting "Attack" and "Release".
386 @code{attack} determines how long the signal has to rise above the threshold
387 before any reduction will occur and @code{release} sets the time the signal
388 has to fall below the threshold to reduce the reduction again. Shorter signals
389 than the chosen attack time will be left untouched.
390 The overall reduction of the signal can be made up afterwards with the
391 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
392 raising the makeup to this level results in a signal twice as loud than the
393 source. To gain a softer entry in the compression the @code{knee} flattens the
394 hard edge at the threshold in the range of the chosen decibels.
396 The filter accepts the following options:
400 Set input gain. Default is 1. Range is between 0.015625 and 64.
403 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
404 Default is @code{downward}.
407 If a signal of stream rises above this level it will affect the gain
409 By default it is 0.125. Range is between 0.00097563 and 1.
412 Set a ratio by which the signal is reduced. 1:2 means that if the level
413 rose 4dB above the threshold, it will be only 2dB above after the reduction.
414 Default is 2. Range is between 1 and 20.
417 Amount of milliseconds the signal has to rise above the threshold before gain
418 reduction starts. Default is 20. Range is between 0.01 and 2000.
421 Amount of milliseconds the signal has to fall below the threshold before
422 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
425 Set the amount by how much signal will be amplified after processing.
426 Default is 1. Range is from 1 to 64.
429 Curve the sharp knee around the threshold to enter gain reduction more softly.
430 Default is 2.82843. Range is between 1 and 8.
433 Choose if the @code{average} level between all channels of input stream
434 or the louder(@code{maximum}) channel of input stream affects the
435 reduction. Default is @code{average}.
438 Should the exact signal be taken in case of @code{peak} or an RMS one in case
439 of @code{rms}. Default is @code{rms} which is mostly smoother.
442 How much to use compressed signal in output. Default is 1.
443 Range is between 0 and 1.
448 This filter supports the all above options as @ref{commands}.
451 Simple audio dynamic range compression/expansion filter.
453 The filter accepts the following options:
457 Set contrast. Default is 33. Allowed range is between 0 and 100.
462 Copy the input audio source unchanged to the output. This is mainly useful for
467 Apply cross fade from one input audio stream to another input audio stream.
468 The cross fade is applied for specified duration near the end of first stream.
470 The filter accepts the following options:
474 Specify the number of samples for which the cross fade effect has to last.
475 At the end of the cross fade effect the first input audio will be completely
476 silent. Default is 44100.
479 Specify the duration of the cross fade effect. See
480 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
481 for the accepted syntax.
482 By default the duration is determined by @var{nb_samples}.
483 If set this option is used instead of @var{nb_samples}.
486 Should first stream end overlap with second stream start. Default is enabled.
489 Set curve for cross fade transition for first stream.
492 Set curve for cross fade transition for second stream.
494 For description of available curve types see @ref{afade} filter description.
501 Cross fade from one input to another:
503 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
507 Cross fade from one input to another but without overlapping:
509 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
514 Split audio stream into several bands.
516 This filter splits audio stream into two or more frequency ranges.
517 Summing all streams back will give flat output.
519 The filter accepts the following options:
523 Set split frequencies. Those must be positive and increasing.
526 Set filter order for each band split. This controls filter roll-off or steepness
527 of filter transfer function.
528 Available values are:
553 Default is @var{4th}.
556 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
559 Set output gain for each band. Default value is 1 for all bands.
566 Split input audio stream into two bands (low and high) with split frequency of 1500 Hz,
567 each band will be in separate stream:
569 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
573 Same as above, but with higher filter order:
575 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500:order=8th[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
579 Same as above, but also with additional middle band (frequencies between 1500 and 8000):
581 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500 8000:order=8th[LOW][MID][HIGH]' -map '[LOW]' low.wav -map '[MID]' mid.wav -map '[HIGH]' high.wav
587 Reduce audio bit resolution.
589 This filter is bit crusher with enhanced functionality. A bit crusher
590 is used to audibly reduce number of bits an audio signal is sampled
591 with. This doesn't change the bit depth at all, it just produces the
592 effect. Material reduced in bit depth sounds more harsh and "digital".
593 This filter is able to even round to continuous values instead of discrete
595 Additionally it has a D/C offset which results in different crushing of
596 the lower and the upper half of the signal.
597 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
599 Another feature of this filter is the logarithmic mode.
600 This setting switches from linear distances between bits to logarithmic ones.
601 The result is a much more "natural" sounding crusher which doesn't gate low
602 signals for example. The human ear has a logarithmic perception,
603 so this kind of crushing is much more pleasant.
604 Logarithmic crushing is also able to get anti-aliased.
606 The filter accepts the following options:
622 Can be linear: @code{lin} or logarithmic: @code{log}.
631 Set sample reduction.
634 Enable LFO. By default disabled.
645 This filter supports the all above options as @ref{commands}.
649 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
653 Remove impulsive noise from input audio.
655 Samples detected as impulsive noise are replaced by interpolated samples using
656 autoregressive modelling.
660 Set window size, in milliseconds. Allowed range is from @code{10} to
661 @code{100}. Default value is @code{55} milliseconds.
662 This sets size of window which will be processed at once.
665 Set window overlap, in percentage of window size. Allowed range is from
666 @code{50} to @code{95}. Default value is @code{75} percent.
667 Setting this to a very high value increases impulsive noise removal but makes
668 whole process much slower.
671 Set autoregression order, in percentage of window size. Allowed range is from
672 @code{0} to @code{25}. Default value is @code{2} percent. This option also
673 controls quality of interpolated samples using neighbour good samples.
676 Set threshold value. Allowed range is from @code{1} to @code{100}.
677 Default value is @code{2}.
678 This controls the strength of impulsive noise which is going to be removed.
679 The lower value, the more samples will be detected as impulsive noise.
682 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
683 @code{10}. Default value is @code{2}.
684 If any two samples detected as noise are spaced less than this value then any
685 sample between those two samples will be also detected as noise.
690 It accepts the following values:
693 Select overlap-add method. Even not interpolated samples are slightly
694 changed with this method.
697 Select overlap-save method. Not interpolated samples remain unchanged.
700 Default value is @code{a}.
704 Remove clipped samples from input audio.
706 Samples detected as clipped are replaced by interpolated samples using
707 autoregressive modelling.
711 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
712 Default value is @code{55} milliseconds.
713 This sets size of window which will be processed at once.
716 Set window overlap, in percentage of window size. Allowed range is from @code{50}
717 to @code{95}. Default value is @code{75} percent.
720 Set autoregression order, in percentage of window size. Allowed range is from
721 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
722 quality of interpolated samples using neighbour good samples.
725 Set threshold value. Allowed range is from @code{1} to @code{100}.
726 Default value is @code{10}. Higher values make clip detection less aggressive.
729 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
730 Default value is @code{1000}. Higher values make clip detection less aggressive.
735 It accepts the following values:
738 Select overlap-add method. Even not interpolated samples are slightly changed
742 Select overlap-save method. Not interpolated samples remain unchanged.
745 Default value is @code{a}.
750 Delay one or more audio channels.
752 Samples in delayed channel are filled with silence.
754 The filter accepts the following option:
758 Set list of delays in milliseconds for each channel separated by '|'.
759 Unused delays will be silently ignored. If number of given delays is
760 smaller than number of channels all remaining channels will not be delayed.
761 If you want to delay exact number of samples, append 'S' to number.
762 If you want instead to delay in seconds, append 's' to number.
765 Use last set delay for all remaining channels. By default is disabled.
766 This option if enabled changes how option @code{delays} is interpreted.
773 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
774 the second channel (and any other channels that may be present) unchanged.
780 Delay second channel by 500 samples, the third channel by 700 samples and leave
781 the first channel (and any other channels that may be present) unchanged.
787 Delay all channels by same number of samples:
789 adelay=delays=64S:all=1
794 Remedy denormals in audio by adding extremely low-level noise.
796 This filter shall be placed before any filter that can produce denormals.
798 A description of the accepted parameters follows.
802 Set level of added noise in dB. Default is @code{-351}.
803 Allowed range is from -451 to -90.
806 Set type of added noise.
819 Default is @code{dc}.
824 This filter supports the all above options as @ref{commands}.
826 @section aderivative, aintegral
828 Compute derivative/integral of audio stream.
830 Applying both filters one after another produces original audio.
834 Apply echoing to the input audio.
836 Echoes are reflected sound and can occur naturally amongst mountains
837 (and sometimes large buildings) when talking or shouting; digital echo
838 effects emulate this behaviour and are often used to help fill out the
839 sound of a single instrument or vocal. The time difference between the
840 original signal and the reflection is the @code{delay}, and the
841 loudness of the reflected signal is the @code{decay}.
842 Multiple echoes can have different delays and decays.
844 A description of the accepted parameters follows.
848 Set input gain of reflected signal. Default is @code{0.6}.
851 Set output gain of reflected signal. Default is @code{0.3}.
854 Set list of time intervals in milliseconds between original signal and reflections
855 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
856 Default is @code{1000}.
859 Set list of loudness of reflected signals separated by '|'.
860 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
861 Default is @code{0.5}.
868 Make it sound as if there are twice as many instruments as are actually playing:
870 aecho=0.8:0.88:60:0.4
874 If delay is very short, then it sounds like a (metallic) robot playing music:
880 A longer delay will sound like an open air concert in the mountains:
882 aecho=0.8:0.9:1000:0.3
886 Same as above but with one more mountain:
888 aecho=0.8:0.9:1000|1800:0.3|0.25
893 Audio emphasis filter creates or restores material directly taken from LPs or
894 emphased CDs with different filter curves. E.g. to store music on vinyl the
895 signal has to be altered by a filter first to even out the disadvantages of
896 this recording medium.
897 Once the material is played back the inverse filter has to be applied to
898 restore the distortion of the frequency response.
900 The filter accepts the following options:
910 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
911 use @code{production} mode. Default is @code{reproduction} mode.
914 Set filter type. Selects medium. Can be one of the following:
926 select Compact Disc (CD).
932 select 50µs (FM-KF).
934 select 75µs (FM-KF).
940 This filter supports the all above options as @ref{commands}.
944 Modify an audio signal according to the specified expressions.
946 This filter accepts one or more expressions (one for each channel),
947 which are evaluated and used to modify a corresponding audio signal.
949 It accepts the following parameters:
953 Set the '|'-separated expressions list for each separate channel. If
954 the number of input channels is greater than the number of
955 expressions, the last specified expression is used for the remaining
958 @item channel_layout, c
959 Set output channel layout. If not specified, the channel layout is
960 specified by the number of expressions. If set to @samp{same}, it will
961 use by default the same input channel layout.
964 Each expression in @var{exprs} can contain the following constants and functions:
968 channel number of the current expression
971 number of the evaluated sample, starting from 0
977 time of the evaluated sample expressed in seconds
980 @item nb_out_channels
981 input and output number of channels
984 the value of input channel with number @var{CH}
987 Note: this filter is slow. For faster processing you should use a
996 aeval=val(ch)/2:c=same
1000 Invert phase of the second channel:
1002 aeval=val(0)|-val(1)
1008 An exciter is used to produce high sound that is not present in the
1009 original signal. This is done by creating harmonic distortions of the
1010 signal which are restricted in range and added to the original signal.
1011 An Exciter raises the upper end of an audio signal without simply raising
1012 the higher frequencies like an equalizer would do to create a more
1013 "crisp" or "brilliant" sound.
1015 The filter accepts the following options:
1019 Set input level prior processing of signal.
1020 Allowed range is from 0 to 64.
1024 Set output level after processing of signal.
1025 Allowed range is from 0 to 64.
1029 Set the amount of harmonics added to original signal.
1030 Allowed range is from 0 to 64.
1034 Set the amount of newly created harmonics.
1035 Allowed range is from 0.1 to 10.
1036 Default value is 8.5.
1039 Set the octave of newly created harmonics.
1040 Allowed range is from -10 to 10.
1044 Set the lower frequency limit of producing harmonics in Hz.
1045 Allowed range is from 2000 to 12000 Hz.
1049 Set the upper frequency limit of producing harmonics.
1050 Allowed range is from 9999 to 20000 Hz.
1051 If value is lower than 10000 Hz no limit is applied.
1054 Mute the original signal and output only added harmonics.
1055 By default is disabled.
1058 @subsection Commands
1060 This filter supports the all above options as @ref{commands}.
1065 Apply fade-in/out effect to input audio.
1067 A description of the accepted parameters follows.
1071 Specify the effect type, can be either @code{in} for fade-in, or
1072 @code{out} for a fade-out effect. Default is @code{in}.
1074 @item start_sample, ss
1075 Specify the number of the start sample for starting to apply the fade
1076 effect. Default is 0.
1078 @item nb_samples, ns
1079 Specify the number of samples for which the fade effect has to last. At
1080 the end of the fade-in effect the output audio will have the same
1081 volume as the input audio, at the end of the fade-out transition
1082 the output audio will be silence. Default is 44100.
1084 @item start_time, st
1085 Specify the start time of the fade effect. Default is 0.
1086 The value must be specified as a time duration; see
1087 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1088 for the accepted syntax.
1089 If set this option is used instead of @var{start_sample}.
1092 Specify the duration of the fade effect. See
1093 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1094 for the accepted syntax.
1095 At the end of the fade-in effect the output audio will have the same
1096 volume as the input audio, at the end of the fade-out transition
1097 the output audio will be silence.
1098 By default the duration is determined by @var{nb_samples}.
1099 If set this option is used instead of @var{nb_samples}.
1102 Set curve for fade transition.
1104 It accepts the following values:
1107 select triangular, linear slope (default)
1109 select quarter of sine wave
1111 select half of sine wave
1113 select exponential sine wave
1117 select inverted parabola
1131 select inverted quarter of sine wave
1133 select inverted half of sine wave
1135 select double-exponential seat
1137 select double-exponential sigmoid
1139 select logistic sigmoid
1141 select sine cardinal function
1143 select inverted sine cardinal function
1149 @subsection Commands
1151 This filter supports the all above options as @ref{commands}.
1153 @subsection Examples
1157 Fade in first 15 seconds of audio:
1159 afade=t=in:ss=0:d=15
1163 Fade out last 25 seconds of a 900 seconds audio:
1165 afade=t=out:st=875:d=25
1170 Denoise audio samples with FFT.
1172 A description of the accepted parameters follows.
1176 Set the noise reduction in dB, allowed range is 0.01 to 97.
1177 Default value is 12 dB.
1180 Set the noise floor in dB, allowed range is -80 to -20.
1181 Default value is -50 dB.
1186 It accepts the following values:
1195 Select shellac noise.
1198 Select custom noise, defined in @code{bn} option.
1200 Default value is white noise.
1204 Set custom band noise for every one of 15 bands.
1205 Bands are separated by ' ' or '|'.
1208 Set the residual floor in dB, allowed range is -80 to -20.
1209 Default value is -38 dB.
1212 Enable noise tracking. By default is disabled.
1213 With this enabled, noise floor is automatically adjusted.
1216 Enable residual tracking. By default is disabled.
1219 Set the output mode.
1221 It accepts the following values:
1224 Pass input unchanged.
1227 Pass noise filtered out.
1232 Default value is @var{o}.
1236 @subsection Commands
1238 This filter supports the following commands:
1240 @item sample_noise, sn
1241 Start or stop measuring noise profile.
1242 Syntax for the command is : "start" or "stop" string.
1243 After measuring noise profile is stopped it will be
1244 automatically applied in filtering.
1246 @item noise_reduction, nr
1247 Change noise reduction. Argument is single float number.
1248 Syntax for the command is : "@var{noise_reduction}"
1250 @item noise_floor, nf
1251 Change noise floor. Argument is single float number.
1252 Syntax for the command is : "@var{noise_floor}"
1254 @item output_mode, om
1255 Change output mode operation.
1256 Syntax for the command is : "i", "o" or "n" string.
1260 Apply arbitrary expressions to samples in frequency domain.
1264 Set frequency domain real expression for each separate channel separated
1265 by '|'. Default is "re".
1266 If the number of input channels is greater than the number of
1267 expressions, the last specified expression is used for the remaining
1271 Set frequency domain imaginary expression for each separate channel
1272 separated by '|'. Default is "im".
1274 Each expression in @var{real} and @var{imag} can contain the following
1275 constants and functions:
1282 current frequency bin number
1285 number of available bins
1288 channel number of the current expression
1297 current real part of frequency bin of current channel
1300 current imaginary part of frequency bin of current channel
1303 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1306 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1310 Set window size. Allowed range is from 16 to 131072.
1311 Default is @code{4096}
1314 Set window function. Default is @code{hann}.
1317 Set window overlap. If set to 1, the recommended overlap for selected
1318 window function will be picked. Default is @code{0.75}.
1321 @subsection Examples
1325 Leave almost only low frequencies in audio:
1327 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1331 Apply robotize effect:
1333 afftfilt="real='hypot(re,im)*sin(0)':imag='hypot(re,im)*cos(0)':win_size=512:overlap=0.75"
1337 Apply whisper effect:
1339 afftfilt="real='hypot(re,im)*cos((random(0)*2-1)*2*3.14)':imag='hypot(re,im)*sin((random(1)*2-1)*2*3.14)':win_size=128:overlap=0.8"
1346 Apply an arbitrary Finite Impulse Response filter.
1348 This filter is designed for applying long FIR filters,
1349 up to 60 seconds long.
1351 It can be used as component for digital crossover filters,
1352 room equalization, cross talk cancellation, wavefield synthesis,
1353 auralization, ambiophonics, ambisonics and spatialization.
1355 This filter uses the streams higher than first one as FIR coefficients.
1356 If the non-first stream holds a single channel, it will be used
1357 for all input channels in the first stream, otherwise
1358 the number of channels in the non-first stream must be same as
1359 the number of channels in the first stream.
1361 It accepts the following parameters:
1365 Set dry gain. This sets input gain.
1368 Set wet gain. This sets final output gain.
1371 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1374 Enable applying gain measured from power of IR.
1376 Set which approach to use for auto gain measurement.
1380 Do not apply any gain.
1383 select peak gain, very conservative approach. This is default value.
1386 select DC gain, limited application.
1389 select gain to noise approach, this is most popular one.
1393 Set gain to be applied to IR coefficients before filtering.
1394 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1397 Set format of IR stream. Can be @code{mono} or @code{input}.
1398 Default is @code{input}.
1401 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1402 Allowed range is 0.1 to 60 seconds.
1405 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1406 By default it is disabled.
1409 Set for which IR channel to display frequency response. By default is first channel
1410 displayed. This option is used only when @var{response} is enabled.
1413 Set video stream size. This option is used only when @var{response} is enabled.
1416 Set video stream frame rate. This option is used only when @var{response} is enabled.
1419 Set minimal partition size used for convolution. Default is @var{8192}.
1420 Allowed range is from @var{1} to @var{32768}.
1421 Lower values decreases latency at cost of higher CPU usage.
1424 Set maximal partition size used for convolution. Default is @var{8192}.
1425 Allowed range is from @var{8} to @var{32768}.
1426 Lower values may increase CPU usage.
1429 Set number of input impulse responses streams which will be switchable at runtime.
1430 Allowed range is from @var{1} to @var{32}. Default is @var{1}.
1433 Set IR stream which will be used for convolution, starting from @var{0}, should always be
1434 lower than supplied value by @code{nbirs} option. Default is @var{0}.
1435 This option can be changed at runtime via @ref{commands}.
1438 @subsection Examples
1442 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1444 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1451 Set output format constraints for the input audio. The framework will
1452 negotiate the most appropriate format to minimize conversions.
1454 It accepts the following parameters:
1457 @item sample_fmts, f
1458 A '|'-separated list of requested sample formats.
1460 @item sample_rates, r
1461 A '|'-separated list of requested sample rates.
1463 @item channel_layouts, cl
1464 A '|'-separated list of requested channel layouts.
1466 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1467 for the required syntax.
1470 If a parameter is omitted, all values are allowed.
1472 Force the output to either unsigned 8-bit or signed 16-bit stereo
1474 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1478 Apply frequency shift to input audio samples.
1480 The filter accepts the following options:
1484 Specify frequency shift. Allowed range is -INT_MAX to INT_MAX.
1485 Default value is 0.0.
1488 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
1489 Default value is 1.0.
1492 @subsection Commands
1494 This filter supports the all above options as @ref{commands}.
1498 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1499 processing reduces disturbing noise between useful signals.
1501 Gating is done by detecting the volume below a chosen level @var{threshold}
1502 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1503 floor is set via @var{range}. Because an exact manipulation of the signal
1504 would cause distortion of the waveform the reduction can be levelled over
1505 time. This is done by setting @var{attack} and @var{release}.
1507 @var{attack} determines how long the signal has to fall below the threshold
1508 before any reduction will occur and @var{release} sets the time the signal
1509 has to rise above the threshold to reduce the reduction again.
1510 Shorter signals than the chosen attack time will be left untouched.
1514 Set input level before filtering.
1515 Default is 1. Allowed range is from 0.015625 to 64.
1518 Set the mode of operation. Can be @code{upward} or @code{downward}.
1519 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1520 will be amplified, expanding dynamic range in upward direction.
1521 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1524 Set the level of gain reduction when the signal is below the threshold.
1525 Default is 0.06125. Allowed range is from 0 to 1.
1526 Setting this to 0 disables reduction and then filter behaves like expander.
1529 If a signal rises above this level the gain reduction is released.
1530 Default is 0.125. Allowed range is from 0 to 1.
1533 Set a ratio by which the signal is reduced.
1534 Default is 2. Allowed range is from 1 to 9000.
1537 Amount of milliseconds the signal has to rise above the threshold before gain
1539 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1542 Amount of milliseconds the signal has to fall below the threshold before the
1543 reduction is increased again. Default is 250 milliseconds.
1544 Allowed range is from 0.01 to 9000.
1547 Set amount of amplification of signal after processing.
1548 Default is 1. Allowed range is from 1 to 64.
1551 Curve the sharp knee around the threshold to enter gain reduction more softly.
1552 Default is 2.828427125. Allowed range is from 1 to 8.
1555 Choose if exact signal should be taken for detection or an RMS like one.
1556 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1559 Choose if the average level between all channels or the louder channel affects
1561 Default is @code{average}. Can be @code{average} or @code{maximum}.
1564 @subsection Commands
1566 This filter supports the all above options as @ref{commands}.
1570 Apply an arbitrary Infinite Impulse Response filter.
1572 It accepts the following parameters:
1576 Set B/numerator/zeros/reflection coefficients.
1579 Set A/denominator/poles/ladder coefficients.
1591 Set coefficients format.
1595 lattice-ladder function
1597 analog transfer function
1599 digital transfer function
1601 Z-plane zeros/poles, cartesian (default)
1603 Z-plane zeros/poles, polar radians
1605 Z-plane zeros/poles, polar degrees
1611 Set type of processing.
1623 Set filtering precision.
1627 double-precision floating-point (default)
1629 single-precision floating-point
1637 Normalize filter coefficients, by default is enabled.
1638 Enabling it will normalize magnitude response at DC to 0dB.
1641 How much to use filtered signal in output. Default is 1.
1642 Range is between 0 and 1.
1645 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1646 By default it is disabled.
1649 Set for which IR channel to display frequency response. By default is first channel
1650 displayed. This option is used only when @var{response} is enabled.
1653 Set video stream size. This option is used only when @var{response} is enabled.
1656 Coefficients in @code{tf} and @code{sf} format are separated by spaces and are in ascending
1659 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1660 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1663 Different coefficients and gains can be provided for every channel, in such case
1664 use '|' to separate coefficients or gains. Last provided coefficients will be
1665 used for all remaining channels.
1667 @subsection Examples
1671 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1673 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1677 Same as above but in @code{zp} format:
1679 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1683 Apply 3-rd order analog normalized Butterworth low-pass filter, using analog transfer function format:
1685 aiir=z=1.3057 0 0 0:p=1.3057 2.3892 2.1860 1:f=sf:r=d
1691 The limiter prevents an input signal from rising over a desired threshold.
1692 This limiter uses lookahead technology to prevent your signal from distorting.
1693 It means that there is a small delay after the signal is processed. Keep in mind
1694 that the delay it produces is the attack time you set.
1696 The filter accepts the following options:
1700 Set input gain. Default is 1.
1703 Set output gain. Default is 1.
1706 Don't let signals above this level pass the limiter. Default is 1.
1709 The limiter will reach its attenuation level in this amount of time in
1710 milliseconds. Default is 5 milliseconds.
1713 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1714 Default is 50 milliseconds.
1717 When gain reduction is always needed ASC takes care of releasing to an
1718 average reduction level rather than reaching a reduction of 0 in the release
1722 Select how much the release time is affected by ASC, 0 means nearly no changes
1723 in release time while 1 produces higher release times.
1726 Auto level output signal. Default is enabled.
1727 This normalizes audio back to 0dB if enabled.
1730 Depending on picked setting it is recommended to upsample input 2x or 4x times
1731 with @ref{aresample} before applying this filter.
1735 Apply a two-pole all-pass filter with central frequency (in Hz)
1736 @var{frequency}, and filter-width @var{width}.
1737 An all-pass filter changes the audio's frequency to phase relationship
1738 without changing its frequency to amplitude relationship.
1740 The filter accepts the following options:
1744 Set frequency in Hz.
1747 Set method to specify band-width of filter.
1762 Specify the band-width of a filter in width_type units.
1765 How much to use filtered signal in output. Default is 1.
1766 Range is between 0 and 1.
1769 Specify which channels to filter, by default all available are filtered.
1772 Normalize biquad coefficients, by default is disabled.
1773 Enabling it will normalize magnitude response at DC to 0dB.
1776 Set the filter order, can be 1 or 2. Default is 2.
1779 Set transform type of IIR filter.
1788 Set precison of filtering.
1791 Pick automatic sample format depending on surround filters.
1793 Always use signed 16-bit.
1795 Always use signed 32-bit.
1797 Always use float 32-bit.
1799 Always use float 64-bit.
1803 @subsection Commands
1805 This filter supports the following commands:
1808 Change allpass frequency.
1809 Syntax for the command is : "@var{frequency}"
1812 Change allpass width_type.
1813 Syntax for the command is : "@var{width_type}"
1816 Change allpass width.
1817 Syntax for the command is : "@var{width}"
1821 Syntax for the command is : "@var{mix}"
1828 The filter accepts the following options:
1832 Set the number of loops. Setting this value to -1 will result in infinite loops.
1836 Set maximal number of samples. Default is 0.
1839 Set first sample of loop. Default is 0.
1845 Merge two or more audio streams into a single multi-channel stream.
1847 The filter accepts the following options:
1852 Set the number of inputs. Default is 2.
1856 If the channel layouts of the inputs are disjoint, and therefore compatible,
1857 the channel layout of the output will be set accordingly and the channels
1858 will be reordered as necessary. If the channel layouts of the inputs are not
1859 disjoint, the output will have all the channels of the first input then all
1860 the channels of the second input, in that order, and the channel layout of
1861 the output will be the default value corresponding to the total number of
1864 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1865 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1866 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1867 first input, b1 is the first channel of the second input).
1869 On the other hand, if both input are in stereo, the output channels will be
1870 in the default order: a1, a2, b1, b2, and the channel layout will be
1871 arbitrarily set to 4.0, which may or may not be the expected value.
1873 All inputs must have the same sample rate, and format.
1875 If inputs do not have the same duration, the output will stop with the
1878 @subsection Examples
1882 Merge two mono files into a stereo stream:
1884 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1888 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1890 ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
1896 Mixes multiple audio inputs into a single output.
1898 Note that this filter only supports float samples (the @var{amerge}
1899 and @var{pan} audio filters support many formats). If the @var{amix}
1900 input has integer samples then @ref{aresample} will be automatically
1901 inserted to perform the conversion to float samples.
1905 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1907 will mix 3 input audio streams to a single output with the same duration as the
1908 first input and a dropout transition time of 3 seconds.
1910 It accepts the following parameters:
1914 The number of inputs. If unspecified, it defaults to 2.
1917 How to determine the end-of-stream.
1921 The duration of the longest input. (default)
1924 The duration of the shortest input.
1927 The duration of the first input.
1931 @item dropout_transition
1932 The transition time, in seconds, for volume renormalization when an input
1933 stream ends. The default value is 2 seconds.
1936 Specify weight of each input audio stream as sequence.
1937 Each weight is separated by space. By default all inputs have same weight.
1940 Always scale inputs instead of only doing summation of samples.
1941 Beware of heavy clipping if inputs are not normalized prior or after filtering
1942 by this filter if this option is disabled. By default is enabled.
1945 @subsection Commands
1947 This filter supports the following commands:
1951 Syntax is same as option with same name.
1956 Multiply first audio stream with second audio stream and store result
1957 in output audio stream. Multiplication is done by multiplying each
1958 sample from first stream with sample at same position from second stream.
1960 With this element-wise multiplication one can create amplitude fades and
1961 amplitude modulations.
1963 @section anequalizer
1965 High-order parametric multiband equalizer for each channel.
1967 It accepts the following parameters:
1971 This option string is in format:
1972 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1973 Each equalizer band is separated by '|'.
1977 Set channel number to which equalization will be applied.
1978 If input doesn't have that channel the entry is ignored.
1981 Set central frequency for band.
1982 If input doesn't have that frequency the entry is ignored.
1985 Set band width in Hertz.
1988 Set band gain in dB.
1991 Set filter type for band, optional, can be:
1995 Butterworth, this is default.
2006 With this option activated frequency response of anequalizer is displayed
2010 Set video stream size. Only useful if curves option is activated.
2013 Set max gain that will be displayed. Only useful if curves option is activated.
2014 Setting this to a reasonable value makes it possible to display gain which is derived from
2015 neighbour bands which are too close to each other and thus produce higher gain
2016 when both are activated.
2019 Set frequency scale used to draw frequency response in video output.
2020 Can be linear or logarithmic. Default is logarithmic.
2023 Set color for each channel curve which is going to be displayed in video stream.
2024 This is list of color names separated by space or by '|'.
2025 Unrecognised or missing colors will be replaced by white color.
2028 @subsection Examples
2032 Lower gain by 10 of central frequency 200Hz and width 100 Hz
2033 for first 2 channels using Chebyshev type 1 filter:
2035 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
2039 @subsection Commands
2041 This filter supports the following commands:
2044 Alter existing filter parameters.
2045 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
2047 @var{fN} is existing filter number, starting from 0, if no such filter is available
2049 @var{freq} set new frequency parameter.
2050 @var{width} set new width parameter in Hertz.
2051 @var{gain} set new gain parameter in dB.
2053 Full filter invocation with asendcmd may look like this:
2054 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
2059 Reduce broadband noise in audio samples using Non-Local Means algorithm.
2061 Each sample is adjusted by looking for other samples with similar contexts. This
2062 context similarity is defined by comparing their surrounding patches of size
2063 @option{p}. Patches are searched in an area of @option{r} around the sample.
2065 The filter accepts the following options:
2069 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
2072 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
2073 Default value is 2 milliseconds.
2076 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
2077 Default value is 6 milliseconds.
2080 Set the output mode.
2082 It accepts the following values:
2085 Pass input unchanged.
2088 Pass noise filtered out.
2093 Default value is @var{o}.
2097 Set smooth factor. Default value is @var{11}. Allowed range is from @var{1} to @var{15}.
2100 @subsection Commands
2102 This filter supports the all above options as @ref{commands}.
2105 Apply Normalized Least-Mean-Squares algorithm to the first audio stream using the second audio stream.
2107 This adaptive filter is used to mimic a desired filter by finding the filter coefficients that
2108 relate to producing the least mean square of the error signal (difference between the desired,
2109 2nd input audio stream and the actual signal, the 1st input audio stream).
2111 A description of the accepted options follows.
2124 Set the filter leakage.
2127 It accepts the following values:
2136 Pass filtered samples.
2139 Pass difference between desired and filtered samples.
2141 Default value is @var{o}.
2145 @subsection Examples
2149 One of many usages of this filter is noise reduction, input audio is filtered
2150 with same samples that are delayed by fixed amount, one such example for stereo audio is:
2152 asplit[a][b],[a]adelay=32S|32S[a],[b][a]anlms=order=128:leakage=0.0005:mu=.5:out_mode=o
2156 @subsection Commands
2158 This filter supports the same commands as options, excluding option @code{order}.
2162 Pass the audio source unchanged to the output.
2166 Pad the end of an audio stream with silence.
2168 This can be used together with @command{ffmpeg} @option{-shortest} to
2169 extend audio streams to the same length as the video stream.
2171 A description of the accepted options follows.
2175 Set silence packet size. Default value is 4096.
2178 Set the number of samples of silence to add to the end. After the
2179 value is reached, the stream is terminated. This option is mutually
2180 exclusive with @option{whole_len}.
2183 Set the minimum total number of samples in the output audio stream. If
2184 the value is longer than the input audio length, silence is added to
2185 the end, until the value is reached. This option is mutually exclusive
2186 with @option{pad_len}.
2189 Specify the duration of samples of silence to add. See
2190 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2191 for the accepted syntax. Used only if set to non-zero value.
2194 Specify the minimum total duration in the output audio stream. See
2195 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2196 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
2197 the input audio length, silence is added to the end, until the value is reached.
2198 This option is mutually exclusive with @option{pad_dur}
2201 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
2202 nor @option{whole_dur} option is set, the filter will add silence to the end of
2203 the input stream indefinitely.
2205 @subsection Examples
2209 Add 1024 samples of silence to the end of the input:
2215 Make sure the audio output will contain at least 10000 samples, pad
2216 the input with silence if required:
2218 apad=whole_len=10000
2222 Use @command{ffmpeg} to pad the audio input with silence, so that the
2223 video stream will always result the shortest and will be converted
2224 until the end in the output file when using the @option{shortest}
2227 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
2232 Add a phasing effect to the input audio.
2234 A phaser filter creates series of peaks and troughs in the frequency spectrum.
2235 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
2237 A description of the accepted parameters follows.
2241 Set input gain. Default is 0.4.
2244 Set output gain. Default is 0.74
2247 Set delay in milliseconds. Default is 3.0.
2250 Set decay. Default is 0.4.
2253 Set modulation speed in Hz. Default is 0.5.
2256 Set modulation type. Default is triangular.
2258 It accepts the following values:
2265 @section aphaseshift
2266 Apply phase shift to input audio samples.
2268 The filter accepts the following options:
2272 Specify phase shift. Allowed range is from -1.0 to 1.0.
2273 Default value is 0.0.
2276 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
2277 Default value is 1.0.
2280 @subsection Commands
2282 This filter supports the all above options as @ref{commands}.
2286 Audio pulsator is something between an autopanner and a tremolo.
2287 But it can produce funny stereo effects as well. Pulsator changes the volume
2288 of the left and right channel based on a LFO (low frequency oscillator) with
2289 different waveforms and shifted phases.
2290 This filter have the ability to define an offset between left and right
2291 channel. An offset of 0 means that both LFO shapes match each other.
2292 The left and right channel are altered equally - a conventional tremolo.
2293 An offset of 50% means that the shape of the right channel is exactly shifted
2294 in phase (or moved backwards about half of the frequency) - pulsator acts as
2295 an autopanner. At 1 both curves match again. Every setting in between moves the
2296 phase shift gapless between all stages and produces some "bypassing" sounds with
2297 sine and triangle waveforms. The more you set the offset near 1 (starting from
2298 the 0.5) the faster the signal passes from the left to the right speaker.
2300 The filter accepts the following options:
2304 Set input gain. By default it is 1. Range is [0.015625 - 64].
2307 Set output gain. By default it is 1. Range is [0.015625 - 64].
2310 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
2311 sawup or sawdown. Default is sine.
2314 Set modulation. Define how much of original signal is affected by the LFO.
2317 Set left channel offset. Default is 0. Allowed range is [0 - 1].
2320 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
2323 Set pulse width. Default is 1. Allowed range is [0 - 2].
2326 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
2329 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
2333 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
2337 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
2338 if timing is set to hz.
2344 Resample the input audio to the specified parameters, using the
2345 libswresample library. If none are specified then the filter will
2346 automatically convert between its input and output.
2348 This filter is also able to stretch/squeeze the audio data to make it match
2349 the timestamps or to inject silence / cut out audio to make it match the
2350 timestamps, do a combination of both or do neither.
2352 The filter accepts the syntax
2353 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
2354 expresses a sample rate and @var{resampler_options} is a list of
2355 @var{key}=@var{value} pairs, separated by ":". See the
2356 @ref{Resampler Options,,"Resampler Options" section in the
2357 ffmpeg-resampler(1) manual,ffmpeg-resampler}
2358 for the complete list of supported options.
2360 @subsection Examples
2364 Resample the input audio to 44100Hz:
2370 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
2371 samples per second compensation:
2373 aresample=async=1000
2379 Reverse an audio clip.
2381 Warning: This filter requires memory to buffer the entire clip, so trimming
2384 @subsection Examples
2388 Take the first 5 seconds of a clip, and reverse it.
2390 atrim=end=5,areverse
2396 Reduce noise from speech using Recurrent Neural Networks.
2398 This filter accepts the following options:
2402 Set train model file to load. This option is always required.
2405 Set how much to mix filtered samples into final output.
2406 Allowed range is from -1 to 1. Default value is 1.
2407 Negative values are special, they set how much to keep filtered noise
2408 in the final filter output. Set this option to -1 to hear actual
2409 noise removed from input signal.
2412 @subsection Commands
2414 This filter supports the all above options as @ref{commands}.
2416 @section asetnsamples
2418 Set the number of samples per each output audio frame.
2420 The last output packet may contain a different number of samples, as
2421 the filter will flush all the remaining samples when the input audio
2424 The filter accepts the following options:
2428 @item nb_out_samples, n
2429 Set the number of frames per each output audio frame. The number is
2430 intended as the number of samples @emph{per each channel}.
2431 Default value is 1024.
2434 If set to 1, the filter will pad the last audio frame with zeroes, so
2435 that the last frame will contain the same number of samples as the
2436 previous ones. Default value is 1.
2439 For example, to set the number of per-frame samples to 1234 and
2440 disable padding for the last frame, use:
2442 asetnsamples=n=1234:p=0
2447 Set the sample rate without altering the PCM data.
2448 This will result in a change of speed and pitch.
2450 The filter accepts the following options:
2453 @item sample_rate, r
2454 Set the output sample rate. Default is 44100 Hz.
2459 Show a line containing various information for each input audio frame.
2460 The input audio is not modified.
2462 The shown line contains a sequence of key/value pairs of the form
2463 @var{key}:@var{value}.
2465 The following values are shown in the output:
2469 The (sequential) number of the input frame, starting from 0.
2472 The presentation timestamp of the input frame, in time base units; the time base
2473 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2476 The presentation timestamp of the input frame in seconds.
2479 position of the frame in the input stream, -1 if this information in
2480 unavailable and/or meaningless (for example in case of synthetic audio)
2489 The sample rate for the audio frame.
2492 The number of samples (per channel) in the frame.
2495 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2496 audio, the data is treated as if all the planes were concatenated.
2498 @item plane_checksums
2499 A list of Adler-32 checksums for each data plane.
2503 Apply audio soft clipping.
2505 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2506 along a smooth curve, rather than the abrupt shape of hard-clipping.
2508 This filter accepts the following options:
2512 Set type of soft-clipping.
2514 It accepts the following values:
2528 Set threshold from where to start clipping. Default value is 0dB or 1.
2531 Set gain applied to output. Default value is 0dB or 1.
2534 Set additional parameter which controls sigmoid function.
2537 Set oversampling factor.
2540 @subsection Commands
2542 This filter supports the all above options as @ref{commands}.
2545 Automatic Speech Recognition
2547 This filter uses PocketSphinx for speech recognition. To enable
2548 compilation of this filter, you need to configure FFmpeg with
2549 @code{--enable-pocketsphinx}.
2551 It accepts the following options:
2555 Set sampling rate of input audio. Defaults is @code{16000}.
2556 This need to match speech models, otherwise one will get poor results.
2559 Set dictionary containing acoustic model files.
2562 Set pronunciation dictionary.
2565 Set language model file.
2568 Set language model set.
2571 Set which language model to use.
2574 Set output for log messages.
2577 The filter exports recognized speech as the frame metadata @code{lavfi.asr.text}.
2582 Display time domain statistical information about the audio channels.
2583 Statistics are calculated and displayed for each audio channel and,
2584 where applicable, an overall figure is also given.
2586 It accepts the following option:
2589 Short window length in seconds, used for peak and trough RMS measurement.
2590 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2594 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2595 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2598 Available keys for each channel are:
2644 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2645 this @code{lavfi.astats.Overall.Peak_count}.
2647 For description what each key means read below.
2650 Set number of frame after which stats are going to be recalculated.
2651 Default is disabled.
2653 @item measure_perchannel
2654 Select the entries which need to be measured per channel. The metadata keys can
2655 be used as flags, default is @option{all} which measures everything.
2656 @option{none} disables all per channel measurement.
2658 @item measure_overall
2659 Select the entries which need to be measured overall. The metadata keys can
2660 be used as flags, default is @option{all} which measures everything.
2661 @option{none} disables all overall measurement.
2665 A description of each shown parameter follows:
2669 Mean amplitude displacement from zero.
2672 Minimal sample level.
2675 Maximal sample level.
2677 @item Min difference
2678 Minimal difference between two consecutive samples.
2680 @item Max difference
2681 Maximal difference between two consecutive samples.
2683 @item Mean difference
2684 Mean difference between two consecutive samples.
2685 The average of each difference between two consecutive samples.
2687 @item RMS difference
2688 Root Mean Square difference between two consecutive samples.
2692 Standard peak and RMS level measured in dBFS.
2696 Peak and trough values for RMS level measured over a short window.
2699 Standard ratio of peak to RMS level (note: not in dB).
2702 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2703 (i.e. either @var{Min level} or @var{Max level}).
2706 Number of occasions (not the number of samples) that the signal attained either
2707 @var{Min level} or @var{Max level}.
2709 @item Noise floor dB
2710 Minimum local peak measured in dBFS over a short window.
2712 @item Noise floor count
2713 Number of occasions (not the number of samples) that the signal attained
2717 Overall bit depth of audio. Number of bits used for each sample.
2720 Measured dynamic range of audio in dB.
2722 @item Zero crossings
2723 Number of points where the waveform crosses the zero level axis.
2725 @item Zero crossings rate
2726 Rate of Zero crossings and number of audio samples.
2730 Boost subwoofer frequencies.
2732 The filter accepts the following options:
2736 Set dry gain, how much of original signal is kept. Allowed range is from 0 to 1.
2737 Default value is 0.7.
2740 Set wet gain, how much of filtered signal is kept. Allowed range is from 0 to 1.
2741 Default value is 0.7.
2744 Set delay line decay gain value. Allowed range is from 0 to 1.
2745 Default value is 0.7.
2748 Set delay line feedback gain value. Allowed range is from 0 to 1.
2749 Default value is 0.9.
2752 Set cutoff frequency in Hertz. Allowed range is 50 to 900.
2753 Default value is 100.
2756 Set slope amount for cutoff frequency. Allowed range is 0.0001 to 1.
2757 Default value is 0.5.
2760 Set delay. Allowed range is from 1 to 100.
2761 Default value is 20.
2764 @subsection Commands
2766 This filter supports the all above options as @ref{commands}.
2769 Cut subwoofer frequencies.
2771 This filter allows to set custom, steeper
2772 roll off than highpass filter, and thus is able to more attenuate
2773 frequency content in stop-band.
2775 The filter accepts the following options:
2779 Set cutoff frequency in Hertz. Allowed range is 2 to 200.
2780 Default value is 20.
2783 Set filter order. Available values are from 3 to 20.
2784 Default value is 10.
2787 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2790 @subsection Commands
2792 This filter supports the all above options as @ref{commands}.
2795 Cut super frequencies.
2797 The filter accepts the following options:
2801 Set cutoff frequency in Hertz. Allowed range is 20000 to 192000.
2802 Default value is 20000.
2805 Set filter order. Available values are from 3 to 20.
2806 Default value is 10.
2809 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2812 @subsection Commands
2814 This filter supports the all above options as @ref{commands}.
2817 Apply high order Butterworth band-pass filter.
2819 The filter accepts the following options:
2823 Set center frequency in Hertz. Allowed range is 2 to 999999.
2824 Default value is 1000.
2827 Set filter order. Available values are from 4 to 20.
2831 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2834 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2837 @subsection Commands
2839 This filter supports the all above options as @ref{commands}.
2842 Apply high order Butterworth band-stop filter.
2844 The filter accepts the following options:
2848 Set center frequency in Hertz. Allowed range is 2 to 999999.
2849 Default value is 1000.
2852 Set filter order. Available values are from 4 to 20.
2856 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2859 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2862 @subsection Commands
2864 This filter supports the all above options as @ref{commands}.
2870 The filter accepts exactly one parameter, the audio tempo. If not
2871 specified then the filter will assume nominal 1.0 tempo. Tempo must
2872 be in the [0.5, 100.0] range.
2874 Note that tempo greater than 2 will skip some samples rather than
2875 blend them in. If for any reason this is a concern it is always
2876 possible to daisy-chain several instances of atempo to achieve the
2877 desired product tempo.
2879 @subsection Examples
2883 Slow down audio to 80% tempo:
2889 To speed up audio to 300% tempo:
2895 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2897 atempo=sqrt(3),atempo=sqrt(3)
2901 @subsection Commands
2903 This filter supports the following commands:
2906 Change filter tempo scale factor.
2907 Syntax for the command is : "@var{tempo}"
2912 Trim the input so that the output contains one continuous subpart of the input.
2914 It accepts the following parameters:
2917 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2918 sample with the timestamp @var{start} will be the first sample in the output.
2921 Specify time of the first audio sample that will be dropped, i.e. the
2922 audio sample immediately preceding the one with the timestamp @var{end} will be
2923 the last sample in the output.
2926 Same as @var{start}, except this option sets the start timestamp in samples
2930 Same as @var{end}, except this option sets the end timestamp in samples instead
2934 The maximum duration of the output in seconds.
2937 The number of the first sample that should be output.
2940 The number of the first sample that should be dropped.
2943 @option{start}, @option{end}, and @option{duration} are expressed as time
2944 duration specifications; see
2945 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2947 Note that the first two sets of the start/end options and the @option{duration}
2948 option look at the frame timestamp, while the _sample options simply count the
2949 samples that pass through the filter. So start/end_pts and start/end_sample will
2950 give different results when the timestamps are wrong, inexact or do not start at
2951 zero. Also note that this filter does not modify the timestamps. If you wish
2952 to have the output timestamps start at zero, insert the asetpts filter after the
2955 If multiple start or end options are set, this filter tries to be greedy and
2956 keep all samples that match at least one of the specified constraints. To keep
2957 only the part that matches all the constraints at once, chain multiple atrim
2960 The defaults are such that all the input is kept. So it is possible to set e.g.
2961 just the end values to keep everything before the specified time.
2966 Drop everything except the second minute of input:
2968 ffmpeg -i INPUT -af atrim=60:120
2972 Keep only the first 1000 samples:
2974 ffmpeg -i INPUT -af atrim=end_sample=1000
2979 @section axcorrelate
2980 Calculate normalized cross-correlation between two input audio streams.
2982 Resulted samples are always between -1 and 1 inclusive.
2983 If result is 1 it means two input samples are highly correlated in that selected segment.
2984 Result 0 means they are not correlated at all.
2985 If result is -1 it means two input samples are out of phase, which means they cancel each
2988 The filter accepts the following options:
2992 Set size of segment over which cross-correlation is calculated.
2993 Default is 256. Allowed range is from 2 to 131072.
2996 Set algorithm for cross-correlation. Can be @code{slow} or @code{fast}.
2997 Default is @code{slow}. Fast algorithm assumes mean values over any given segment
2998 are always zero and thus need much less calculations to make.
2999 This is generally not true, but is valid for typical audio streams.
3002 @subsection Examples
3006 Calculate correlation between channels in stereo audio stream:
3008 ffmpeg -i stereo.wav -af channelsplit,axcorrelate=size=1024:algo=fast correlation.wav
3014 Apply a two-pole Butterworth band-pass filter with central
3015 frequency @var{frequency}, and (3dB-point) band-width width.
3016 The @var{csg} option selects a constant skirt gain (peak gain = Q)
3017 instead of the default: constant 0dB peak gain.
3018 The filter roll off at 6dB per octave (20dB per decade).
3020 The filter accepts the following options:
3024 Set the filter's central frequency. Default is @code{3000}.
3027 Constant skirt gain if set to 1. Defaults to 0.
3030 Set method to specify band-width of filter.
3045 Specify the band-width of a filter in width_type units.
3048 How much to use filtered signal in output. Default is 1.
3049 Range is between 0 and 1.
3052 Specify which channels to filter, by default all available are filtered.
3055 Normalize biquad coefficients, by default is disabled.
3056 Enabling it will normalize magnitude response at DC to 0dB.
3059 Set transform type of IIR filter.
3068 Set precison of filtering.
3071 Pick automatic sample format depending on surround filters.
3073 Always use signed 16-bit.
3075 Always use signed 32-bit.
3077 Always use float 32-bit.
3079 Always use float 64-bit.
3083 @subsection Commands
3085 This filter supports the following commands:
3088 Change bandpass frequency.
3089 Syntax for the command is : "@var{frequency}"
3092 Change bandpass width_type.
3093 Syntax for the command is : "@var{width_type}"
3096 Change bandpass width.
3097 Syntax for the command is : "@var{width}"
3100 Change bandpass mix.
3101 Syntax for the command is : "@var{mix}"
3106 Apply a two-pole Butterworth band-reject filter with central
3107 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
3108 The filter roll off at 6dB per octave (20dB per decade).
3110 The filter accepts the following options:
3114 Set the filter's central frequency. Default is @code{3000}.
3117 Set method to specify band-width of filter.
3132 Specify the band-width of a filter in width_type units.
3135 How much to use filtered signal in output. Default is 1.
3136 Range is between 0 and 1.
3139 Specify which channels to filter, by default all available are filtered.
3142 Normalize biquad coefficients, by default is disabled.
3143 Enabling it will normalize magnitude response at DC to 0dB.
3146 Set transform type of IIR filter.
3155 Set precison of filtering.
3158 Pick automatic sample format depending on surround filters.
3160 Always use signed 16-bit.
3162 Always use signed 32-bit.
3164 Always use float 32-bit.
3166 Always use float 64-bit.
3170 @subsection Commands
3172 This filter supports the following commands:
3175 Change bandreject frequency.
3176 Syntax for the command is : "@var{frequency}"
3179 Change bandreject width_type.
3180 Syntax for the command is : "@var{width_type}"
3183 Change bandreject width.
3184 Syntax for the command is : "@var{width}"
3187 Change bandreject mix.
3188 Syntax for the command is : "@var{mix}"
3191 @section bass, lowshelf
3193 Boost or cut the bass (lower) frequencies of the audio using a two-pole
3194 shelving filter with a response similar to that of a standard
3195 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3197 The filter accepts the following options:
3201 Give the gain at 0 Hz. Its useful range is about -20
3202 (for a large cut) to +20 (for a large boost).
3203 Beware of clipping when using a positive gain.
3206 Set the filter's central frequency and so can be used
3207 to extend or reduce the frequency range to be boosted or cut.
3208 The default value is @code{100} Hz.
3211 Set method to specify band-width of filter.
3226 Determine how steep is the filter's shelf transition.
3229 Set number of poles. Default is 2.
3232 How much to use filtered signal in output. Default is 1.
3233 Range is between 0 and 1.
3236 Specify which channels to filter, by default all available are filtered.
3239 Normalize biquad coefficients, by default is disabled.
3240 Enabling it will normalize magnitude response at DC to 0dB.
3243 Set transform type of IIR filter.
3252 Set precison of filtering.
3255 Pick automatic sample format depending on surround filters.
3257 Always use signed 16-bit.
3259 Always use signed 32-bit.
3261 Always use float 32-bit.
3263 Always use float 64-bit.
3267 @subsection Commands
3269 This filter supports the following commands:
3272 Change bass frequency.
3273 Syntax for the command is : "@var{frequency}"
3276 Change bass width_type.
3277 Syntax for the command is : "@var{width_type}"
3281 Syntax for the command is : "@var{width}"
3285 Syntax for the command is : "@var{gain}"
3289 Syntax for the command is : "@var{mix}"
3294 Apply a biquad IIR filter with the given coefficients.
3295 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
3296 are the numerator and denominator coefficients respectively.
3297 and @var{channels}, @var{c} specify which channels to filter, by default all
3298 available are filtered.
3300 @subsection Commands
3302 This filter supports the following commands:
3310 Change biquad parameter.
3311 Syntax for the command is : "@var{value}"
3314 How much to use filtered signal in output. Default is 1.
3315 Range is between 0 and 1.
3318 Specify which channels to filter, by default all available are filtered.
3321 Normalize biquad coefficients, by default is disabled.
3322 Enabling it will normalize magnitude response at DC to 0dB.
3325 Set transform type of IIR filter.
3334 Set precison of filtering.
3337 Pick automatic sample format depending on surround filters.
3339 Always use signed 16-bit.
3341 Always use signed 32-bit.
3343 Always use float 32-bit.
3345 Always use float 64-bit.
3350 Bauer stereo to binaural transformation, which improves headphone listening of
3351 stereo audio records.
3353 To enable compilation of this filter you need to configure FFmpeg with
3354 @code{--enable-libbs2b}.
3356 It accepts the following parameters:
3360 Pre-defined crossfeed level.
3364 Default level (fcut=700, feed=50).
3367 Chu Moy circuit (fcut=700, feed=60).
3370 Jan Meier circuit (fcut=650, feed=95).
3375 Cut frequency (in Hz).
3384 Remap input channels to new locations.
3386 It accepts the following parameters:
3389 Map channels from input to output. The argument is a '|'-separated list of
3390 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
3391 @var{in_channel} form. @var{in_channel} can be either the name of the input
3392 channel (e.g. FL for front left) or its index in the input channel layout.
3393 @var{out_channel} is the name of the output channel or its index in the output
3394 channel layout. If @var{out_channel} is not given then it is implicitly an
3395 index, starting with zero and increasing by one for each mapping.
3397 @item channel_layout
3398 The channel layout of the output stream.
3401 If no mapping is present, the filter will implicitly map input channels to
3402 output channels, preserving indices.
3404 @subsection Examples
3408 For example, assuming a 5.1+downmix input MOV file,
3410 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
3412 will create an output WAV file tagged as stereo from the downmix channels of
3416 To fix a 5.1 WAV improperly encoded in AAC's native channel order
3418 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
3422 @section channelsplit
3424 Split each channel from an input audio stream into a separate output stream.
3426 It accepts the following parameters:
3428 @item channel_layout
3429 The channel layout of the input stream. The default is "stereo".
3431 A channel layout describing the channels to be extracted as separate output streams
3432 or "all" to extract each input channel as a separate stream. The default is "all".
3434 Choosing channels not present in channel layout in the input will result in an error.
3437 @subsection Examples
3441 For example, assuming a stereo input MP3 file,
3443 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
3445 will create an output Matroska file with two audio streams, one containing only
3446 the left channel and the other the right channel.
3449 Split a 5.1 WAV file into per-channel files:
3451 ffmpeg -i in.wav -filter_complex
3452 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
3453 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
3454 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
3459 Extract only LFE from a 5.1 WAV file:
3461 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
3462 -map '[LFE]' lfe.wav
3467 Add a chorus effect to the audio.
3469 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
3471 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
3472 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
3473 The modulation depth defines the range the modulated delay is played before or after
3474 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
3475 sound tuned around the original one, like in a chorus where some vocals are slightly
3478 It accepts the following parameters:
3481 Set input gain. Default is 0.4.
3484 Set output gain. Default is 0.4.
3487 Set delays. A typical delay is around 40ms to 60ms.
3499 @subsection Examples
3505 chorus=0.7:0.9:55:0.4:0.25:2
3511 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
3515 Fuller sounding chorus with three delays:
3517 chorus=0.5:0.9:50|60|40:0.4|0.32|0.3:0.25|0.4|0.3:2|2.3|1.3
3522 Compress or expand the audio's dynamic range.
3524 It accepts the following parameters:
3530 A list of times in seconds for each channel over which the instantaneous level
3531 of the input signal is averaged to determine its volume. @var{attacks} refers to
3532 increase of volume and @var{decays} refers to decrease of volume. For most
3533 situations, the attack time (response to the audio getting louder) should be
3534 shorter than the decay time, because the human ear is more sensitive to sudden
3535 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
3536 a typical value for decay is 0.8 seconds.
3537 If specified number of attacks & decays is lower than number of channels, the last
3538 set attack/decay will be used for all remaining channels.
3541 A list of points for the transfer function, specified in dB relative to the
3542 maximum possible signal amplitude. Each key points list must be defined using
3543 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
3544 @code{x0/y0 x1/y1 x2/y2 ....}
3546 The input values must be in strictly increasing order but the transfer function
3547 does not have to be monotonically rising. The point @code{0/0} is assumed but
3548 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
3549 function are @code{-70/-70|-60/-20|1/0}.
3552 Set the curve radius in dB for all joints. It defaults to 0.01.
3555 Set the additional gain in dB to be applied at all points on the transfer
3556 function. This allows for easy adjustment of the overall gain.
3560 Set an initial volume, in dB, to be assumed for each channel when filtering
3561 starts. This permits the user to supply a nominal level initially, so that, for
3562 example, a very large gain is not applied to initial signal levels before the
3563 companding has begun to operate. A typical value for audio which is initially
3564 quiet is -90 dB. It defaults to 0.
3567 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
3568 delayed before being fed to the volume adjuster. Specifying a delay
3569 approximately equal to the attack/decay times allows the filter to effectively
3570 operate in predictive rather than reactive mode. It defaults to 0.
3574 @subsection Examples
3578 Make music with both quiet and loud passages suitable for listening to in a
3581 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
3584 Another example for audio with whisper and explosion parts:
3586 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
3590 A noise gate for when the noise is at a lower level than the signal:
3592 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
3596 Here is another noise gate, this time for when the noise is at a higher level
3597 than the signal (making it, in some ways, similar to squelch):
3599 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
3603 2:1 compression starting at -6dB:
3605 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
3609 2:1 compression starting at -9dB:
3611 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
3615 2:1 compression starting at -12dB:
3617 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
3621 2:1 compression starting at -18dB:
3623 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
3627 3:1 compression starting at -15dB:
3629 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
3635 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
3641 compand=attacks=0:points=-80/-169|-54/-80|-49.5/-64.6|-41.1/-41.1|-25.8/-15|-10.8/-4.5|0/0|20/8.3
3645 Hard limiter at -6dB:
3647 compand=attacks=0:points=-80/-80|-6/-6|20/-6
3651 Hard limiter at -12dB:
3653 compand=attacks=0:points=-80/-80|-12/-12|20/-12
3657 Hard noise gate at -35 dB:
3659 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
3665 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
3669 @section compensationdelay
3671 Compensation Delay Line is a metric based delay to compensate differing
3672 positions of microphones or speakers.
3674 For example, you have recorded guitar with two microphones placed in
3675 different locations. Because the front of sound wave has fixed speed in
3676 normal conditions, the phasing of microphones can vary and depends on
3677 their location and interposition. The best sound mix can be achieved when
3678 these microphones are in phase (synchronized). Note that a distance of
3679 ~30 cm between microphones makes one microphone capture the signal in
3680 antiphase to the other microphone. That makes the final mix sound moody.
3681 This filter helps to solve phasing problems by adding different delays
3682 to each microphone track and make them synchronized.
3684 The best result can be reached when you take one track as base and
3685 synchronize other tracks one by one with it.
3686 Remember that synchronization/delay tolerance depends on sample rate, too.
3687 Higher sample rates will give more tolerance.
3689 The filter accepts the following parameters:
3693 Set millimeters distance. This is compensation distance for fine tuning.
3697 Set cm distance. This is compensation distance for tightening distance setup.
3701 Set meters distance. This is compensation distance for hard distance setup.
3705 Set dry amount. Amount of unprocessed (dry) signal.
3709 Set wet amount. Amount of processed (wet) signal.
3713 Set temperature in degrees Celsius. This is the temperature of the environment.
3718 Apply headphone crossfeed filter.
3720 Crossfeed is the process of blending the left and right channels of stereo
3722 It is mainly used to reduce extreme stereo separation of low frequencies.
3724 The intent is to produce more speaker like sound to the listener.
3726 The filter accepts the following options:
3730 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
3731 This sets gain of low shelf filter for side part of stereo image.
3732 Default is -6dB. Max allowed is -30db when strength is set to 1.
3735 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
3736 This sets cut off frequency of low shelf filter. Default is cut off near
3737 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
3740 Set curve slope of low shelf filter. Default is 0.5.
3741 Allowed range is from 0.01 to 1.
3744 Set input gain. Default is 0.9.
3747 Set output gain. Default is 1.
3750 @subsection Commands
3752 This filter supports the all above options as @ref{commands}.
3754 @section crystalizer
3755 Simple algorithm for audio noise sharpening.
3757 This filter linearly increases differences betweeen each audio sample.
3759 The filter accepts the following options:
3763 Sets the intensity of effect (default: 2.0). Must be in range between -10.0 to 0
3764 (unchanged sound) to 10.0 (maximum effect).
3765 To inverse filtering use negative value.
3768 Enable clipping. By default is enabled.
3771 @subsection Commands
3773 This filter supports the all above options as @ref{commands}.
3776 Apply a DC shift to the audio.
3778 This can be useful to remove a DC offset (caused perhaps by a hardware problem
3779 in the recording chain) from the audio. The effect of a DC offset is reduced
3780 headroom and hence volume. The @ref{astats} filter can be used to determine if
3781 a signal has a DC offset.
3785 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
3789 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
3790 used to prevent clipping.
3795 Apply de-essing to the audio samples.
3799 Set intensity for triggering de-essing. Allowed range is from 0 to 1.
3803 Set amount of ducking on treble part of sound. Allowed range is from 0 to 1.
3807 How much of original frequency content to keep when de-essing. Allowed range is from 0 to 1.
3811 Set the output mode.
3813 It accepts the following values:
3816 Pass input unchanged.
3819 Pass ess filtered out.
3824 Default value is @var{o}.
3830 Measure audio dynamic range.
3832 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3833 is found in transition material. And anything less that 8 have very poor dynamics
3834 and is very compressed.
3836 The filter accepts the following options:
3840 Set window length in seconds used to split audio into segments of equal length.
3841 Default is 3 seconds.
3845 Dynamic Audio Normalizer.
3847 This filter applies a certain amount of gain to the input audio in order
3848 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3849 contrast to more "simple" normalization algorithms, the Dynamic Audio
3850 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3851 This allows for applying extra gain to the "quiet" sections of the audio
3852 while avoiding distortions or clipping the "loud" sections. In other words:
3853 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3854 sections, in the sense that the volume of each section is brought to the
3855 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3856 this goal *without* applying "dynamic range compressing". It will retain 100%
3857 of the dynamic range *within* each section of the audio file.
3861 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3862 Default is 500 milliseconds.
3863 The Dynamic Audio Normalizer processes the input audio in small chunks,
3864 referred to as frames. This is required, because a peak magnitude has no
3865 meaning for just a single sample value. Instead, we need to determine the
3866 peak magnitude for a contiguous sequence of sample values. While a "standard"
3867 normalizer would simply use the peak magnitude of the complete file, the
3868 Dynamic Audio Normalizer determines the peak magnitude individually for each
3869 frame. The length of a frame is specified in milliseconds. By default, the
3870 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3871 been found to give good results with most files.
3872 Note that the exact frame length, in number of samples, will be determined
3873 automatically, based on the sampling rate of the individual input audio file.
3876 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3877 number. Default is 31.
3878 Probably the most important parameter of the Dynamic Audio Normalizer is the
3879 @code{window size} of the Gaussian smoothing filter. The filter's window size
3880 is specified in frames, centered around the current frame. For the sake of
3881 simplicity, this must be an odd number. Consequently, the default value of 31
3882 takes into account the current frame, as well as the 15 preceding frames and
3883 the 15 subsequent frames. Using a larger window results in a stronger
3884 smoothing effect and thus in less gain variation, i.e. slower gain
3885 adaptation. Conversely, using a smaller window results in a weaker smoothing
3886 effect and thus in more gain variation, i.e. faster gain adaptation.
3887 In other words, the more you increase this value, the more the Dynamic Audio
3888 Normalizer will behave like a "traditional" normalization filter. On the
3889 contrary, the more you decrease this value, the more the Dynamic Audio
3890 Normalizer will behave like a dynamic range compressor.
3893 Set the target peak value. This specifies the highest permissible magnitude
3894 level for the normalized audio input. This filter will try to approach the
3895 target peak magnitude as closely as possible, but at the same time it also
3896 makes sure that the normalized signal will never exceed the peak magnitude.
3897 A frame's maximum local gain factor is imposed directly by the target peak
3898 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3899 It is not recommended to go above this value.
3902 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3903 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3904 factor for each input frame, i.e. the maximum gain factor that does not
3905 result in clipping or distortion. The maximum gain factor is determined by
3906 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3907 additionally bounds the frame's maximum gain factor by a predetermined
3908 (global) maximum gain factor. This is done in order to avoid excessive gain
3909 factors in "silent" or almost silent frames. By default, the maximum gain
3910 factor is 10.0, For most inputs the default value should be sufficient and
3911 it usually is not recommended to increase this value. Though, for input
3912 with an extremely low overall volume level, it may be necessary to allow even
3913 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3914 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3915 Instead, a "sigmoid" threshold function will be applied. This way, the
3916 gain factors will smoothly approach the threshold value, but never exceed that
3920 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3921 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3922 This means that the maximum local gain factor for each frame is defined
3923 (only) by the frame's highest magnitude sample. This way, the samples can
3924 be amplified as much as possible without exceeding the maximum signal
3925 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3926 Normalizer can also take into account the frame's root mean square,
3927 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3928 determine the power of a time-varying signal. It is therefore considered
3929 that the RMS is a better approximation of the "perceived loudness" than
3930 just looking at the signal's peak magnitude. Consequently, by adjusting all
3931 frames to a constant RMS value, a uniform "perceived loudness" can be
3932 established. If a target RMS value has been specified, a frame's local gain
3933 factor is defined as the factor that would result in exactly that RMS value.
3934 Note, however, that the maximum local gain factor is still restricted by the
3935 frame's highest magnitude sample, in order to prevent clipping.
3938 Enable channels coupling. By default is enabled.
3939 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3940 amount. This means the same gain factor will be applied to all channels, i.e.
3941 the maximum possible gain factor is determined by the "loudest" channel.
3942 However, in some recordings, it may happen that the volume of the different
3943 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3944 In this case, this option can be used to disable the channel coupling. This way,
3945 the gain factor will be determined independently for each channel, depending
3946 only on the individual channel's highest magnitude sample. This allows for
3947 harmonizing the volume of the different channels.
3950 Enable DC bias correction. By default is disabled.
3951 An audio signal (in the time domain) is a sequence of sample values.
3952 In the Dynamic Audio Normalizer these sample values are represented in the
3953 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3954 audio signal, or "waveform", should be centered around the zero point.
3955 That means if we calculate the mean value of all samples in a file, or in a
3956 single frame, then the result should be 0.0 or at least very close to that
3957 value. If, however, there is a significant deviation of the mean value from
3958 0.0, in either positive or negative direction, this is referred to as a
3959 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3960 Audio Normalizer provides optional DC bias correction.
3961 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3962 the mean value, or "DC correction" offset, of each input frame and subtract
3963 that value from all of the frame's sample values which ensures those samples
3964 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3965 boundaries, the DC correction offset values will be interpolated smoothly
3966 between neighbouring frames.
3968 @item altboundary, b
3969 Enable alternative boundary mode. By default is disabled.
3970 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3971 around each frame. This includes the preceding frames as well as the
3972 subsequent frames. However, for the "boundary" frames, located at the very
3973 beginning and at the very end of the audio file, not all neighbouring
3974 frames are available. In particular, for the first few frames in the audio
3975 file, the preceding frames are not known. And, similarly, for the last few
3976 frames in the audio file, the subsequent frames are not known. Thus, the
3977 question arises which gain factors should be assumed for the missing frames
3978 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3979 to deal with this situation. The default boundary mode assumes a gain factor
3980 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3981 "fade out" at the beginning and at the end of the input, respectively.
3984 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3985 By default, the Dynamic Audio Normalizer does not apply "traditional"
3986 compression. This means that signal peaks will not be pruned and thus the
3987 full dynamic range will be retained within each local neighbourhood. However,
3988 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3989 normalization algorithm with a more "traditional" compression.
3990 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3991 (thresholding) function. If (and only if) the compression feature is enabled,
3992 all input frames will be processed by a soft knee thresholding function prior
3993 to the actual normalization process. Put simply, the thresholding function is
3994 going to prune all samples whose magnitude exceeds a certain threshold value.
3995 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3996 value. Instead, the threshold value will be adjusted for each individual
3998 In general, smaller parameters result in stronger compression, and vice versa.
3999 Values below 3.0 are not recommended, because audible distortion may appear.
4002 Set the target threshold value. This specifies the lowest permissible
4003 magnitude level for the audio input which will be normalized.
4004 If input frame volume is above this value frame will be normalized.
4005 Otherwise frame may not be normalized at all. The default value is set
4006 to 0, which means all input frames will be normalized.
4007 This option is mostly useful if digital noise is not wanted to be amplified.
4010 @subsection Commands
4012 This filter supports the all above options as @ref{commands}.
4016 Make audio easier to listen to on headphones.
4018 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
4019 so that when listened to on headphones the stereo image is moved from
4020 inside your head (standard for headphones) to outside and in front of
4021 the listener (standard for speakers).
4027 Apply a two-pole peaking equalisation (EQ) filter. With this
4028 filter, the signal-level at and around a selected frequency can
4029 be increased or decreased, whilst (unlike bandpass and bandreject
4030 filters) that at all other frequencies is unchanged.
4032 In order to produce complex equalisation curves, this filter can
4033 be given several times, each with a different central frequency.
4035 The filter accepts the following options:
4039 Set the filter's central frequency in Hz.
4042 Set method to specify band-width of filter.
4057 Specify the band-width of a filter in width_type units.
4060 Set the required gain or attenuation in dB.
4061 Beware of clipping when using a positive gain.
4064 How much to use filtered signal in output. Default is 1.
4065 Range is between 0 and 1.
4068 Specify which channels to filter, by default all available are filtered.
4071 Normalize biquad coefficients, by default is disabled.
4072 Enabling it will normalize magnitude response at DC to 0dB.
4075 Set transform type of IIR filter.
4084 Set precison of filtering.
4087 Pick automatic sample format depending on surround filters.
4089 Always use signed 16-bit.
4091 Always use signed 32-bit.
4093 Always use float 32-bit.
4095 Always use float 64-bit.
4099 @subsection Examples
4102 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
4104 equalizer=f=1000:t=h:width=200:g=-10
4108 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
4110 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
4114 @subsection Commands
4116 This filter supports the following commands:
4119 Change equalizer frequency.
4120 Syntax for the command is : "@var{frequency}"
4123 Change equalizer width_type.
4124 Syntax for the command is : "@var{width_type}"
4127 Change equalizer width.
4128 Syntax for the command is : "@var{width}"
4131 Change equalizer gain.
4132 Syntax for the command is : "@var{gain}"
4135 Change equalizer mix.
4136 Syntax for the command is : "@var{mix}"
4139 @section extrastereo
4141 Linearly increases the difference between left and right channels which
4142 adds some sort of "live" effect to playback.
4144 The filter accepts the following options:
4148 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
4149 (average of both channels), with 1.0 sound will be unchanged, with
4150 -1.0 left and right channels will be swapped.
4153 Enable clipping. By default is enabled.
4156 @subsection Commands
4158 This filter supports the all above options as @ref{commands}.
4160 @section firequalizer
4161 Apply FIR Equalization using arbitrary frequency response.
4163 The filter accepts the following option:
4167 Set gain curve equation (in dB). The expression can contain variables:
4170 the evaluated frequency
4174 channel number, set to 0 when multichannels evaluation is disabled
4176 channel id, see libavutil/channel_layout.h, set to the first channel id when
4177 multichannels evaluation is disabled
4181 channel_layout, see libavutil/channel_layout.h
4186 @item gain_interpolate(f)
4187 interpolate gain on frequency f based on gain_entry
4188 @item cubic_interpolate(f)
4189 same as gain_interpolate, but smoother
4191 This option is also available as command. Default is @code{gain_interpolate(f)}.
4194 Set gain entry for gain_interpolate function. The expression can
4198 store gain entry at frequency f with value g
4200 This option is also available as command.
4203 Set filter delay in seconds. Higher value means more accurate.
4204 Default is @code{0.01}.
4207 Set filter accuracy in Hz. Lower value means more accurate.
4208 Default is @code{5}.
4211 Set window function. Acceptable values are:
4214 rectangular window, useful when gain curve is already smooth
4216 hann window (default)
4222 3-terms continuous 1st derivative nuttall window
4224 minimum 3-terms discontinuous nuttall window
4226 4-terms continuous 1st derivative nuttall window
4228 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
4230 blackman-harris window
4236 If enabled, use fixed number of audio samples. This improves speed when
4237 filtering with large delay. Default is disabled.
4240 Enable multichannels evaluation on gain. Default is disabled.
4243 Enable zero phase mode by subtracting timestamp to compensate delay.
4244 Default is disabled.
4247 Set scale used by gain. Acceptable values are:
4250 linear frequency, linear gain
4252 linear frequency, logarithmic (in dB) gain (default)
4254 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
4256 logarithmic frequency, logarithmic gain
4260 Set file for dumping, suitable for gnuplot.
4263 Set scale for dumpfile. Acceptable values are same with scale option.
4267 Enable 2-channel convolution using complex FFT. This improves speed significantly.
4268 Default is disabled.
4271 Enable minimum phase impulse response. Default is disabled.
4274 @subsection Examples
4279 firequalizer=gain='if(lt(f,1000), 0, -INF)'
4282 lowpass at 1000 Hz with gain_entry:
4284 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
4287 custom equalization:
4289 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
4292 higher delay with zero phase to compensate delay:
4294 firequalizer=delay=0.1:fixed=on:zero_phase=on
4297 lowpass on left channel, highpass on right channel:
4299 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
4300 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
4305 Apply a flanging effect to the audio.
4307 The filter accepts the following options:
4311 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
4314 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
4317 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
4321 Set percentage of delayed signal mixed with original. Range from 0 to 100.
4322 Default value is 71.
4325 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
4328 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
4329 Default value is @var{sinusoidal}.
4332 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
4333 Default value is 25.
4336 Set delay-line interpolation, @var{linear} or @var{quadratic}.
4337 Default is @var{linear}.
4341 Apply Haas effect to audio.
4343 Note that this makes most sense to apply on mono signals.
4344 With this filter applied to mono signals it give some directionality and
4345 stretches its stereo image.
4347 The filter accepts the following options:
4351 Set input level. By default is @var{1}, or 0dB
4354 Set output level. By default is @var{1}, or 0dB.
4357 Set gain applied to side part of signal. By default is @var{1}.
4360 Set kind of middle source. Can be one of the following:
4370 Pick middle part signal of stereo image.
4373 Pick side part signal of stereo image.
4377 Change middle phase. By default is disabled.
4380 Set left channel delay. By default is @var{2.05} milliseconds.
4383 Set left channel balance. By default is @var{-1}.
4386 Set left channel gain. By default is @var{1}.
4389 Change left phase. By default is disabled.
4392 Set right channel delay. By defaults is @var{2.12} milliseconds.
4395 Set right channel balance. By default is @var{1}.
4398 Set right channel gain. By default is @var{1}.
4401 Change right phase. By default is enabled.
4406 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
4407 embedded HDCD codes is expanded into a 20-bit PCM stream.
4409 The filter supports the Peak Extend and Low-level Gain Adjustment features
4410 of HDCD, and detects the Transient Filter flag.
4413 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
4416 When using the filter with wav, note the default encoding for wav is 16-bit,
4417 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
4418 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
4420 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
4421 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
4424 The filter accepts the following options:
4427 @item disable_autoconvert
4428 Disable any automatic format conversion or resampling in the filter graph.
4430 @item process_stereo
4431 Process the stereo channels together. If target_gain does not match between
4432 channels, consider it invalid and use the last valid target_gain.
4435 Set the code detect timer period in ms.
4438 Always extend peaks above -3dBFS even if PE isn't signaled.
4441 Replace audio with a solid tone and adjust the amplitude to signal some
4442 specific aspect of the decoding process. The output file can be loaded in
4443 an audio editor alongside the original to aid analysis.
4445 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
4452 Gain adjustment level at each sample
4454 Samples where peak extend occurs
4456 Samples where the code detect timer is active
4458 Samples where the target gain does not match between channels
4464 Apply head-related transfer functions (HRTFs) to create virtual
4465 loudspeakers around the user for binaural listening via headphones.
4466 The HRIRs are provided via additional streams, for each channel
4467 one stereo input stream is needed.
4469 The filter accepts the following options:
4473 Set mapping of input streams for convolution.
4474 The argument is a '|'-separated list of channel names in order as they
4475 are given as additional stream inputs for filter.
4476 This also specify number of input streams. Number of input streams
4477 must be not less than number of channels in first stream plus one.
4480 Set gain applied to audio. Value is in dB. Default is 0.
4483 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4484 processing audio in time domain which is slow.
4485 @var{freq} is processing audio in frequency domain which is fast.
4486 Default is @var{freq}.
4489 Set custom gain for LFE channels. Value is in dB. Default is 0.
4492 Set size of frame in number of samples which will be processed at once.
4493 Default value is @var{1024}. Allowed range is from 1024 to 96000.
4496 Set format of hrir stream.
4497 Default value is @var{stereo}. Alternative value is @var{multich}.
4498 If value is set to @var{stereo}, number of additional streams should
4499 be greater or equal to number of input channels in first input stream.
4500 Also each additional stream should have stereo number of channels.
4501 If value is set to @var{multich}, number of additional streams should
4502 be exactly one. Also number of input channels of additional stream
4503 should be equal or greater than twice number of channels of first input
4507 @subsection Examples
4511 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4512 each amovie filter use stereo file with IR coefficients as input.
4513 The files give coefficients for each position of virtual loudspeaker:
4516 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
4521 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4522 but now in @var{multich} @var{hrir} format.
4524 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
4531 Apply a high-pass filter with 3dB point frequency.
4532 The filter can be either single-pole, or double-pole (the default).
4533 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4535 The filter accepts the following options:
4539 Set frequency in Hz. Default is 3000.
4542 Set number of poles. Default is 2.
4545 Set method to specify band-width of filter.
4560 Specify the band-width of a filter in width_type units.
4561 Applies only to double-pole filter.
4562 The default is 0.707q and gives a Butterworth response.
4565 How much to use filtered signal in output. Default is 1.
4566 Range is between 0 and 1.
4569 Specify which channels to filter, by default all available are filtered.
4572 Normalize biquad coefficients, by default is disabled.
4573 Enabling it will normalize magnitude response at DC to 0dB.
4576 Set transform type of IIR filter.
4585 Set precison of filtering.
4588 Pick automatic sample format depending on surround filters.
4590 Always use signed 16-bit.
4592 Always use signed 32-bit.
4594 Always use float 32-bit.
4596 Always use float 64-bit.
4600 @subsection Commands
4602 This filter supports the following commands:
4605 Change highpass frequency.
4606 Syntax for the command is : "@var{frequency}"
4609 Change highpass width_type.
4610 Syntax for the command is : "@var{width_type}"
4613 Change highpass width.
4614 Syntax for the command is : "@var{width}"
4617 Change highpass mix.
4618 Syntax for the command is : "@var{mix}"
4623 Join multiple input streams into one multi-channel stream.
4625 It accepts the following parameters:
4629 The number of input streams. It defaults to 2.
4631 @item channel_layout
4632 The desired output channel layout. It defaults to stereo.
4635 Map channels from inputs to output. The argument is a '|'-separated list of
4636 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
4637 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
4638 can be either the name of the input channel (e.g. FL for front left) or its
4639 index in the specified input stream. @var{out_channel} is the name of the output
4643 The filter will attempt to guess the mappings when they are not specified
4644 explicitly. It does so by first trying to find an unused matching input channel
4645 and if that fails it picks the first unused input channel.
4647 Join 3 inputs (with properly set channel layouts):
4649 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
4652 Build a 5.1 output from 6 single-channel streams:
4654 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
4655 'join=inputs=6:channel_layout=5.1:map=0.0-FL|1.0-FR|2.0-FC|3.0-SL|4.0-SR|5.0-LFE'
4661 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
4663 To enable compilation of this filter you need to configure FFmpeg with
4664 @code{--enable-ladspa}.
4668 Specifies the name of LADSPA plugin library to load. If the environment
4669 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
4670 each one of the directories specified by the colon separated list in
4671 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
4672 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
4673 @file{/usr/lib/ladspa/}.
4676 Specifies the plugin within the library. Some libraries contain only
4677 one plugin, but others contain many of them. If this is not set filter
4678 will list all available plugins within the specified library.
4681 Set the '|' separated list of controls which are zero or more floating point
4682 values that determine the behavior of the loaded plugin (for example delay,
4684 Controls need to be defined using the following syntax:
4685 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
4686 @var{valuei} is the value set on the @var{i}-th control.
4687 Alternatively they can be also defined using the following syntax:
4688 @var{value0}|@var{value1}|@var{value2}|..., where
4689 @var{valuei} is the value set on the @var{i}-th control.
4690 If @option{controls} is set to @code{help}, all available controls and
4691 their valid ranges are printed.
4693 @item sample_rate, s
4694 Specify the sample rate, default to 44100. Only used if plugin have
4698 Set the number of samples per channel per each output frame, default
4699 is 1024. Only used if plugin have zero inputs.
4702 Set the minimum duration of the sourced audio. See
4703 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4704 for the accepted syntax.
4705 Note that the resulting duration may be greater than the specified duration,
4706 as the generated audio is always cut at the end of a complete frame.
4707 If not specified, or the expressed duration is negative, the audio is
4708 supposed to be generated forever.
4709 Only used if plugin have zero inputs.
4712 Enable latency compensation, by default is disabled.
4713 Only used if plugin have inputs.
4716 @subsection Examples
4720 List all available plugins within amp (LADSPA example plugin) library:
4726 List all available controls and their valid ranges for @code{vcf_notch}
4727 plugin from @code{VCF} library:
4729 ladspa=f=vcf:p=vcf_notch:c=help
4733 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
4736 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
4740 Add reverberation to the audio using TAP-plugins
4741 (Tom's Audio Processing plugins):
4743 ladspa=file=tap_reverb:tap_reverb
4747 Generate white noise, with 0.2 amplitude:
4749 ladspa=file=cmt:noise_source_white:c=c0=.2
4753 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
4754 @code{C* Audio Plugin Suite} (CAPS) library:
4756 ladspa=file=caps:Click:c=c1=20'
4760 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
4762 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
4766 Increase volume by 20dB using fast lookahead limiter from Steve Harris
4767 @code{SWH Plugins} collection:
4769 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
4773 Attenuate low frequencies using Multiband EQ from Steve Harris
4774 @code{SWH Plugins} collection:
4776 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
4780 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
4783 ladspa=caps:Narrower
4787 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
4789 ladspa=caps:White:.2
4793 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
4795 ladspa=caps:Fractal:c=c1=1
4799 Dynamic volume normalization using @code{VLevel} plugin:
4801 ladspa=vlevel-ladspa:vlevel_mono
4805 @subsection Commands
4807 This filter supports the following commands:
4810 Modify the @var{N}-th control value.
4812 If the specified value is not valid, it is ignored and prior one is kept.
4817 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
4818 Support for both single pass (livestreams, files) and double pass (files) modes.
4819 This algorithm can target IL, LRA, and maximum true peak. In dynamic mode, to accurately
4820 detect true peaks, the audio stream will be upsampled to 192 kHz.
4821 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
4823 The filter accepts the following options:
4827 Set integrated loudness target.
4828 Range is -70.0 - -5.0. Default value is -24.0.
4831 Set loudness range target.
4832 Range is 1.0 - 20.0. Default value is 7.0.
4835 Set maximum true peak.
4836 Range is -9.0 - +0.0. Default value is -2.0.
4838 @item measured_I, measured_i
4839 Measured IL of input file.
4840 Range is -99.0 - +0.0.
4842 @item measured_LRA, measured_lra
4843 Measured LRA of input file.
4844 Range is 0.0 - 99.0.
4846 @item measured_TP, measured_tp
4847 Measured true peak of input file.
4848 Range is -99.0 - +99.0.
4850 @item measured_thresh
4851 Measured threshold of input file.
4852 Range is -99.0 - +0.0.
4855 Set offset gain. Gain is applied before the true-peak limiter.
4856 Range is -99.0 - +99.0. Default is +0.0.
4859 Normalize by linearly scaling the source audio.
4860 @code{measured_I}, @code{measured_LRA}, @code{measured_TP},
4861 and @code{measured_thresh} must all be specified. Target LRA shouldn't
4862 be lower than source LRA and the change in integrated loudness shouldn't
4863 result in a true peak which exceeds the target TP. If any of these
4864 conditions aren't met, normalization mode will revert to @var{dynamic}.
4865 Options are @code{true} or @code{false}. Default is @code{true}.
4868 Treat mono input files as "dual-mono". If a mono file is intended for playback
4869 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
4870 If set to @code{true}, this option will compensate for this effect.
4871 Multi-channel input files are not affected by this option.
4872 Options are true or false. Default is false.
4875 Set print format for stats. Options are summary, json, or none.
4876 Default value is none.
4881 Apply a low-pass filter with 3dB point frequency.
4882 The filter can be either single-pole or double-pole (the default).
4883 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4885 The filter accepts the following options:
4889 Set frequency in Hz. Default is 500.
4892 Set number of poles. Default is 2.
4895 Set method to specify band-width of filter.
4910 Specify the band-width of a filter in width_type units.
4911 Applies only to double-pole filter.
4912 The default is 0.707q and gives a Butterworth response.
4915 How much to use filtered signal in output. Default is 1.
4916 Range is between 0 and 1.
4919 Specify which channels to filter, by default all available are filtered.
4922 Normalize biquad coefficients, by default is disabled.
4923 Enabling it will normalize magnitude response at DC to 0dB.
4926 Set transform type of IIR filter.
4935 Set precison of filtering.
4938 Pick automatic sample format depending on surround filters.
4940 Always use signed 16-bit.
4942 Always use signed 32-bit.
4944 Always use float 32-bit.
4946 Always use float 64-bit.
4950 @subsection Examples
4953 Lowpass only LFE channel, it LFE is not present it does nothing:
4959 @subsection Commands
4961 This filter supports the following commands:
4964 Change lowpass frequency.
4965 Syntax for the command is : "@var{frequency}"
4968 Change lowpass width_type.
4969 Syntax for the command is : "@var{width_type}"
4972 Change lowpass width.
4973 Syntax for the command is : "@var{width}"
4977 Syntax for the command is : "@var{mix}"
4982 Load a LV2 (LADSPA Version 2) plugin.
4984 To enable compilation of this filter you need to configure FFmpeg with
4985 @code{--enable-lv2}.
4989 Specifies the plugin URI. You may need to escape ':'.
4992 Set the '|' separated list of controls which are zero or more floating point
4993 values that determine the behavior of the loaded plugin (for example delay,
4995 If @option{controls} is set to @code{help}, all available controls and
4996 their valid ranges are printed.
4998 @item sample_rate, s
4999 Specify the sample rate, default to 44100. Only used if plugin have
5003 Set the number of samples per channel per each output frame, default
5004 is 1024. Only used if plugin have zero inputs.
5007 Set the minimum duration of the sourced audio. See
5008 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5009 for the accepted syntax.
5010 Note that the resulting duration may be greater than the specified duration,
5011 as the generated audio is always cut at the end of a complete frame.
5012 If not specified, or the expressed duration is negative, the audio is
5013 supposed to be generated forever.
5014 Only used if plugin have zero inputs.
5017 @subsection Examples
5021 Apply bass enhancer plugin from Calf:
5023 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
5027 Apply vinyl plugin from Calf:
5029 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
5033 Apply bit crusher plugin from ArtyFX:
5035 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
5040 Multiband Compress or expand the audio's dynamic range.
5042 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
5043 This is akin to the crossover of a loudspeaker, and results in flat frequency
5044 response when absent compander action.
5046 It accepts the following parameters:
5050 This option syntax is:
5051 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
5052 For explanation of each item refer to compand filter documentation.
5058 Mix channels with specific gain levels. The filter accepts the output
5059 channel layout followed by a set of channels definitions.
5061 This filter is also designed to efficiently remap the channels of an audio
5064 The filter accepts parameters of the form:
5065 "@var{l}|@var{outdef}|@var{outdef}|..."
5069 output channel layout or number of channels
5072 output channel specification, of the form:
5073 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
5076 output channel to define, either a channel name (FL, FR, etc.) or a channel
5077 number (c0, c1, etc.)
5080 multiplicative coefficient for the channel, 1 leaving the volume unchanged
5083 input channel to use, see out_name for details; it is not possible to mix
5084 named and numbered input channels
5087 If the `=' in a channel specification is replaced by `<', then the gains for
5088 that specification will be renormalized so that the total is 1, thus
5089 avoiding clipping noise.
5091 @subsection Mixing examples
5093 For example, if you want to down-mix from stereo to mono, but with a bigger
5094 factor for the left channel:
5096 pan=1c|c0=0.9*c0+0.1*c1
5099 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
5100 7-channels surround:
5102 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
5105 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
5106 that should be preferred (see "-ac" option) unless you have very specific
5109 @subsection Remapping examples
5111 The channel remapping will be effective if, and only if:
5114 @item gain coefficients are zeroes or ones,
5115 @item only one input per channel output,
5118 If all these conditions are satisfied, the filter will notify the user ("Pure
5119 channel mapping detected"), and use an optimized and lossless method to do the
5122 For example, if you have a 5.1 source and want a stereo audio stream by
5123 dropping the extra channels:
5125 pan="stereo| c0=FL | c1=FR"
5128 Given the same source, you can also switch front left and front right channels
5129 and keep the input channel layout:
5131 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
5134 If the input is a stereo audio stream, you can mute the front left channel (and
5135 still keep the stereo channel layout) with:
5140 Still with a stereo audio stream input, you can copy the right channel in both
5141 front left and right:
5143 pan="stereo| c0=FR | c1=FR"
5148 ReplayGain scanner filter. This filter takes an audio stream as an input and
5149 outputs it unchanged.
5150 At end of filtering it displays @code{track_gain} and @code{track_peak}.
5154 Convert the audio sample format, sample rate and channel layout. It is
5155 not meant to be used directly.
5158 Apply time-stretching and pitch-shifting with librubberband.
5160 To enable compilation of this filter, you need to configure FFmpeg with
5161 @code{--enable-librubberband}.
5163 The filter accepts the following options:
5167 Set tempo scale factor.
5170 Set pitch scale factor.
5173 Set transients detector.
5174 Possible values are:
5183 Possible values are:
5192 Possible values are:
5199 Set processing window size.
5200 Possible values are:
5209 Possible values are:
5216 Enable formant preservation when shift pitching.
5217 Possible values are:
5225 Possible values are:
5234 Possible values are:
5241 @subsection Commands
5243 This filter supports the following commands:
5246 Change filter tempo scale factor.
5247 Syntax for the command is : "@var{tempo}"
5250 Change filter pitch scale factor.
5251 Syntax for the command is : "@var{pitch}"
5254 @section sidechaincompress
5256 This filter acts like normal compressor but has the ability to compress
5257 detected signal using second input signal.
5258 It needs two input streams and returns one output stream.
5259 First input stream will be processed depending on second stream signal.
5260 The filtered signal then can be filtered with other filters in later stages of
5261 processing. See @ref{pan} and @ref{amerge} filter.
5263 The filter accepts the following options:
5267 Set input gain. Default is 1. Range is between 0.015625 and 64.
5270 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
5271 Default is @code{downward}.
5274 If a signal of second stream raises above this level it will affect the gain
5275 reduction of first stream.
5276 By default is 0.125. Range is between 0.00097563 and 1.
5279 Set a ratio about which the signal is reduced. 1:2 means that if the level
5280 raised 4dB above the threshold, it will be only 2dB above after the reduction.
5281 Default is 2. Range is between 1 and 20.
5284 Amount of milliseconds the signal has to rise above the threshold before gain
5285 reduction starts. Default is 20. Range is between 0.01 and 2000.
5288 Amount of milliseconds the signal has to fall below the threshold before
5289 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
5292 Set the amount by how much signal will be amplified after processing.
5293 Default is 1. Range is from 1 to 64.
5296 Curve the sharp knee around the threshold to enter gain reduction more softly.
5297 Default is 2.82843. Range is between 1 and 8.
5300 Choose if the @code{average} level between all channels of side-chain stream
5301 or the louder(@code{maximum}) channel of side-chain stream affects the
5302 reduction. Default is @code{average}.
5305 Should the exact signal be taken in case of @code{peak} or an RMS one in case
5306 of @code{rms}. Default is @code{rms} which is mainly smoother.
5309 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
5312 How much to use compressed signal in output. Default is 1.
5313 Range is between 0 and 1.
5316 @subsection Commands
5318 This filter supports the all above options as @ref{commands}.
5320 @subsection Examples
5324 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
5325 depending on the signal of 2nd input and later compressed signal to be
5326 merged with 2nd input:
5328 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
5332 @section sidechaingate
5334 A sidechain gate acts like a normal (wideband) gate but has the ability to
5335 filter the detected signal before sending it to the gain reduction stage.
5336 Normally a gate uses the full range signal to detect a level above the
5338 For example: If you cut all lower frequencies from your sidechain signal
5339 the gate will decrease the volume of your track only if not enough highs
5340 appear. With this technique you are able to reduce the resonation of a
5341 natural drum or remove "rumbling" of muted strokes from a heavily distorted
5343 It needs two input streams and returns one output stream.
5344 First input stream will be processed depending on second stream signal.
5346 The filter accepts the following options:
5350 Set input level before filtering.
5351 Default is 1. Allowed range is from 0.015625 to 64.
5354 Set the mode of operation. Can be @code{upward} or @code{downward}.
5355 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
5356 will be amplified, expanding dynamic range in upward direction.
5357 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
5360 Set the level of gain reduction when the signal is below the threshold.
5361 Default is 0.06125. Allowed range is from 0 to 1.
5362 Setting this to 0 disables reduction and then filter behaves like expander.
5365 If a signal rises above this level the gain reduction is released.
5366 Default is 0.125. Allowed range is from 0 to 1.
5369 Set a ratio about which the signal is reduced.
5370 Default is 2. Allowed range is from 1 to 9000.
5373 Amount of milliseconds the signal has to rise above the threshold before gain
5375 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
5378 Amount of milliseconds the signal has to fall below the threshold before the
5379 reduction is increased again. Default is 250 milliseconds.
5380 Allowed range is from 0.01 to 9000.
5383 Set amount of amplification of signal after processing.
5384 Default is 1. Allowed range is from 1 to 64.
5387 Curve the sharp knee around the threshold to enter gain reduction more softly.
5388 Default is 2.828427125. Allowed range is from 1 to 8.
5391 Choose if exact signal should be taken for detection or an RMS like one.
5392 Default is rms. Can be peak or rms.
5395 Choose if the average level between all channels or the louder channel affects
5397 Default is average. Can be average or maximum.
5400 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
5403 @subsection Commands
5405 This filter supports the all above options as @ref{commands}.
5407 @section silencedetect
5409 Detect silence in an audio stream.
5411 This filter logs a message when it detects that the input audio volume is less
5412 or equal to a noise tolerance value for a duration greater or equal to the
5413 minimum detected noise duration.
5415 The printed times and duration are expressed in seconds. The
5416 @code{lavfi.silence_start} or @code{lavfi.silence_start.X} metadata key
5417 is set on the first frame whose timestamp equals or exceeds the detection
5418 duration and it contains the timestamp of the first frame of the silence.
5420 The @code{lavfi.silence_duration} or @code{lavfi.silence_duration.X}
5421 and @code{lavfi.silence_end} or @code{lavfi.silence_end.X} metadata
5422 keys are set on the first frame after the silence. If @option{mono} is
5423 enabled, and each channel is evaluated separately, the @code{.X}
5424 suffixed keys are used, and @code{X} corresponds to the channel number.
5426 The filter accepts the following options:
5430 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
5431 specified value) or amplitude ratio. Default is -60dB, or 0.001.
5434 Set silence duration until notification (default is 2 seconds). See
5435 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5436 for the accepted syntax.
5439 Process each channel separately, instead of combined. By default is disabled.
5442 @subsection Examples
5446 Detect 5 seconds of silence with -50dB noise tolerance:
5448 silencedetect=n=-50dB:d=5
5452 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
5453 tolerance in @file{silence.mp3}:
5455 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
5459 @section silenceremove
5461 Remove silence from the beginning, middle or end of the audio.
5463 The filter accepts the following options:
5467 This value is used to indicate if audio should be trimmed at beginning of
5468 the audio. A value of zero indicates no silence should be trimmed from the
5469 beginning. When specifying a non-zero value, it trims audio up until it
5470 finds non-silence. Normally, when trimming silence from beginning of audio
5471 the @var{start_periods} will be @code{1} but it can be increased to higher
5472 values to trim all audio up to specific count of non-silence periods.
5473 Default value is @code{0}.
5475 @item start_duration
5476 Specify the amount of time that non-silence must be detected before it stops
5477 trimming audio. By increasing the duration, bursts of noises can be treated
5478 as silence and trimmed off. Default value is @code{0}.
5480 @item start_threshold
5481 This indicates what sample value should be treated as silence. For digital
5482 audio, a value of @code{0} may be fine but for audio recorded from analog,
5483 you may wish to increase the value to account for background noise.
5484 Can be specified in dB (in case "dB" is appended to the specified value)
5485 or amplitude ratio. Default value is @code{0}.
5488 Specify max duration of silence at beginning that will be kept after
5489 trimming. Default is 0, which is equal to trimming all samples detected
5493 Specify mode of detection of silence end in start of multi-channel audio.
5494 Can be @var{any} or @var{all}. Default is @var{any}.
5495 With @var{any}, any sample that is detected as non-silence will cause
5496 stopped trimming of silence.
5497 With @var{all}, only if all channels are detected as non-silence will cause
5498 stopped trimming of silence.
5501 Set the count for trimming silence from the end of audio.
5502 To remove silence from the middle of a file, specify a @var{stop_periods}
5503 that is negative. This value is then treated as a positive value and is
5504 used to indicate the effect should restart processing as specified by
5505 @var{start_periods}, making it suitable for removing periods of silence
5506 in the middle of the audio.
5507 Default value is @code{0}.
5510 Specify a duration of silence that must exist before audio is not copied any
5511 more. By specifying a higher duration, silence that is wanted can be left in
5513 Default value is @code{0}.
5515 @item stop_threshold
5516 This is the same as @option{start_threshold} but for trimming silence from
5518 Can be specified in dB (in case "dB" is appended to the specified value)
5519 or amplitude ratio. Default value is @code{0}.
5522 Specify max duration of silence at end that will be kept after
5523 trimming. Default is 0, which is equal to trimming all samples detected
5527 Specify mode of detection of silence start in end of multi-channel audio.
5528 Can be @var{any} or @var{all}. Default is @var{any}.
5529 With @var{any}, any sample that is detected as non-silence will cause
5530 stopped trimming of silence.
5531 With @var{all}, only if all channels are detected as non-silence will cause
5532 stopped trimming of silence.
5535 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
5536 and works better with digital silence which is exactly 0.
5537 Default value is @code{rms}.
5540 Set duration in number of seconds used to calculate size of window in number
5541 of samples for detecting silence.
5542 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
5545 @subsection Examples
5549 The following example shows how this filter can be used to start a recording
5550 that does not contain the delay at the start which usually occurs between
5551 pressing the record button and the start of the performance:
5553 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
5557 Trim all silence encountered from beginning to end where there is more than 1
5558 second of silence in audio:
5560 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
5564 Trim all digital silence samples, using peak detection, from beginning to end
5565 where there is more than 0 samples of digital silence in audio and digital
5566 silence is detected in all channels at same positions in stream:
5568 silenceremove=window=0:detection=peak:stop_mode=all:start_mode=all:stop_periods=-1:stop_threshold=0
5574 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
5575 loudspeakers around the user for binaural listening via headphones (audio
5576 formats up to 9 channels supported).
5577 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
5578 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
5579 Austrian Academy of Sciences.
5581 To enable compilation of this filter you need to configure FFmpeg with
5582 @code{--enable-libmysofa}.
5584 The filter accepts the following options:
5588 Set the SOFA file used for rendering.
5591 Set gain applied to audio. Value is in dB. Default is 0.
5594 Set rotation of virtual loudspeakers in deg. Default is 0.
5597 Set elevation of virtual speakers in deg. Default is 0.
5600 Set distance in meters between loudspeakers and the listener with near-field
5601 HRTFs. Default is 1.
5604 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
5605 processing audio in time domain which is slow.
5606 @var{freq} is processing audio in frequency domain which is fast.
5607 Default is @var{freq}.
5610 Set custom positions of virtual loudspeakers. Syntax for this option is:
5611 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
5612 Each virtual loudspeaker is described with short channel name following with
5613 azimuth and elevation in degrees.
5614 Each virtual loudspeaker description is separated by '|'.
5615 For example to override front left and front right channel positions use:
5616 'speakers=FL 45 15|FR 345 15'.
5617 Descriptions with unrecognised channel names are ignored.
5620 Set custom gain for LFE channels. Value is in dB. Default is 0.
5623 Set custom frame size in number of samples. Default is 1024.
5624 Allowed range is from 1024 to 96000. Only used if option @samp{type}
5625 is set to @var{freq}.
5628 Should all IRs be normalized upon importing SOFA file.
5629 By default is enabled.
5632 Should nearest IRs be interpolated with neighbor IRs if exact position
5633 does not match. By default is disabled.
5636 Minphase all IRs upon loading of SOFA file. By default is disabled.
5639 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
5642 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
5645 @subsection Examples
5649 Using ClubFritz6 sofa file:
5651 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
5655 Using ClubFritz12 sofa file and bigger radius with small rotation:
5657 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
5661 Similar as above but with custom speaker positions for front left, front right, back left and back right
5662 and also with custom gain:
5664 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
5671 This filter expands or compresses each half-cycle of audio samples
5672 (local set of samples all above or all below zero and between two nearest zero crossings) depending
5673 on threshold value, so audio reaches target peak value under conditions controlled by below options.
5675 The filter accepts the following options:
5679 Set the expansion target peak value. This specifies the highest allowed absolute amplitude
5680 level for the normalized audio input. Default value is 0.95. Allowed range is from 0.0 to 1.0.
5683 Set the maximum expansion factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5684 This option controls maximum local half-cycle of samples expansion. The maximum expansion
5685 would be such that local peak value reaches target peak value but never to surpass it and that
5686 ratio between new and previous peak value does not surpass this option value.
5688 @item compression, c
5689 Set the maximum compression factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5690 This option controls maximum local half-cycle of samples compression. This option is used
5691 only if @option{threshold} option is set to value greater than 0.0, then in such cases
5692 when local peak is lower or same as value set by @option{threshold} all samples belonging to
5693 that peak's half-cycle will be compressed by current compression factor.
5696 Set the threshold value. Default value is 0.0. Allowed range is from 0.0 to 1.0.
5697 This option specifies which half-cycles of samples will be compressed and which will be expanded.
5698 Any half-cycle samples with their local peak value below or same as this option value will be
5699 compressed by current compression factor, otherwise, if greater than threshold value they will be
5700 expanded with expansion factor so that it could reach peak target value but never surpass it.
5703 Set the expansion raising amount per each half-cycle of samples. Default value is 0.001.
5704 Allowed range is from 0.0 to 1.0. This controls how fast expansion factor is raised per
5705 each new half-cycle until it reaches @option{expansion} value.
5706 Setting this options too high may lead to distortions.
5709 Set the compression raising amount per each half-cycle of samples. Default value is 0.001.
5710 Allowed range is from 0.0 to 1.0. This controls how fast compression factor is raised per
5711 each new half-cycle until it reaches @option{compression} value.
5714 Specify which channels to filter, by default all available channels are filtered.
5717 Enable inverted filtering, by default is disabled. This inverts interpretation of @option{threshold}
5718 option. When enabled any half-cycle of samples with their local peak value below or same as
5719 @option{threshold} option will be expanded otherwise it will be compressed.
5722 Link channels when calculating gain applied to each filtered channel sample, by default is disabled.
5723 When disabled each filtered channel gain calculation is independent, otherwise when this option
5724 is enabled the minimum of all possible gains for each filtered channel is used.
5727 @subsection Commands
5729 This filter supports the all above options as @ref{commands}.
5731 @section stereotools
5733 This filter has some handy utilities to manage stereo signals, for converting
5734 M/S stereo recordings to L/R signal while having control over the parameters
5735 or spreading the stereo image of master track.
5737 The filter accepts the following options:
5741 Set input level before filtering for both channels. Defaults is 1.
5742 Allowed range is from 0.015625 to 64.
5745 Set output level after filtering for both channels. Defaults is 1.
5746 Allowed range is from 0.015625 to 64.
5749 Set input balance between both channels. Default is 0.
5750 Allowed range is from -1 to 1.
5753 Set output balance between both channels. Default is 0.
5754 Allowed range is from -1 to 1.
5757 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
5758 clipping. Disabled by default.
5761 Mute the left channel. Disabled by default.
5764 Mute the right channel. Disabled by default.
5767 Change the phase of the left channel. Disabled by default.
5770 Change the phase of the right channel. Disabled by default.
5773 Set stereo mode. Available values are:
5777 Left/Right to Left/Right, this is default.
5780 Left/Right to Mid/Side.
5783 Mid/Side to Left/Right.
5786 Left/Right to Left/Left.
5789 Left/Right to Right/Right.
5792 Left/Right to Left + Right.
5795 Left/Right to Right/Left.
5798 Mid/Side to Left/Left.
5801 Mid/Side to Right/Right.
5804 Mid/Side to Right/Left.
5807 Left/Right to Left - Right.
5811 Set level of side signal. Default is 1.
5812 Allowed range is from 0.015625 to 64.
5815 Set balance of side signal. Default is 0.
5816 Allowed range is from -1 to 1.
5819 Set level of the middle signal. Default is 1.
5820 Allowed range is from 0.015625 to 64.
5823 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
5826 Set stereo base between mono and inversed channels. Default is 0.
5827 Allowed range is from -1 to 1.
5830 Set delay in milliseconds how much to delay left from right channel and
5831 vice versa. Default is 0. Allowed range is from -20 to 20.
5834 Set S/C level. Default is 1. Allowed range is from 1 to 100.
5837 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
5839 @item bmode_in, bmode_out
5840 Set balance mode for balance_in/balance_out option.
5842 Can be one of the following:
5846 Classic balance mode. Attenuate one channel at time.
5847 Gain is raised up to 1.
5850 Similar as classic mode above but gain is raised up to 2.
5853 Equal power distribution, from -6dB to +6dB range.
5857 @subsection Commands
5859 This filter supports the all above options as @ref{commands}.
5861 @subsection Examples
5865 Apply karaoke like effect:
5867 stereotools=mlev=0.015625
5871 Convert M/S signal to L/R:
5873 "stereotools=mode=ms>lr"
5877 @section stereowiden
5879 This filter enhance the stereo effect by suppressing signal common to both
5880 channels and by delaying the signal of left into right and vice versa,
5881 thereby widening the stereo effect.
5883 The filter accepts the following options:
5887 Time in milliseconds of the delay of left signal into right and vice versa.
5888 Default is 20 milliseconds.
5891 Amount of gain in delayed signal into right and vice versa. Gives a delay
5892 effect of left signal in right output and vice versa which gives widening
5893 effect. Default is 0.3.
5896 Cross feed of left into right with inverted phase. This helps in suppressing
5897 the mono. If the value is 1 it will cancel all the signal common to both
5898 channels. Default is 0.3.
5901 Set level of input signal of original channel. Default is 0.8.
5904 @subsection Commands
5906 This filter supports the all above options except @code{delay} as @ref{commands}.
5908 @section superequalizer
5909 Apply 18 band equalizer.
5911 The filter accepts the following options:
5918 Set 131Hz band gain.
5920 Set 185Hz band gain.
5922 Set 262Hz band gain.
5924 Set 370Hz band gain.
5926 Set 523Hz band gain.
5928 Set 740Hz band gain.
5930 Set 1047Hz band gain.
5932 Set 1480Hz band gain.
5934 Set 2093Hz band gain.
5936 Set 2960Hz band gain.
5938 Set 4186Hz band gain.
5940 Set 5920Hz band gain.
5942 Set 8372Hz band gain.
5944 Set 11840Hz band gain.
5946 Set 16744Hz band gain.
5948 Set 20000Hz band gain.
5952 Apply audio surround upmix filter.
5954 This filter allows to produce multichannel output from audio stream.
5956 The filter accepts the following options:
5960 Set output channel layout. By default, this is @var{5.1}.
5962 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5963 for the required syntax.
5966 Set input channel layout. By default, this is @var{stereo}.
5968 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5969 for the required syntax.
5972 Set input volume level. By default, this is @var{1}.
5975 Set output volume level. By default, this is @var{1}.
5978 Enable LFE channel output if output channel layout has it. By default, this is enabled.
5981 Set LFE low cut off frequency. By default, this is @var{128} Hz.
5984 Set LFE high cut off frequency. By default, this is @var{256} Hz.
5987 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
5988 In @var{add} mode, LFE channel is created from input audio and added to output.
5989 In @var{sub} mode, LFE channel is created from input audio and added to output but
5990 also all non-LFE output channels are subtracted with output LFE channel.
5993 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
5994 Default is @var{90}.
5997 Set front center input volume. By default, this is @var{1}.
6000 Set front center output volume. By default, this is @var{1}.
6003 Set front left input volume. By default, this is @var{1}.
6006 Set front left output volume. By default, this is @var{1}.
6009 Set front right input volume. By default, this is @var{1}.
6012 Set front right output volume. By default, this is @var{1}.
6015 Set side left input volume. By default, this is @var{1}.
6018 Set side left output volume. By default, this is @var{1}.
6021 Set side right input volume. By default, this is @var{1}.
6024 Set side right output volume. By default, this is @var{1}.
6027 Set back left input volume. By default, this is @var{1}.
6030 Set back left output volume. By default, this is @var{1}.
6033 Set back right input volume. By default, this is @var{1}.
6036 Set back right output volume. By default, this is @var{1}.
6039 Set back center input volume. By default, this is @var{1}.
6042 Set back center output volume. By default, this is @var{1}.
6045 Set LFE input volume. By default, this is @var{1}.
6048 Set LFE output volume. By default, this is @var{1}.
6051 Set spread usage of stereo image across X axis for all channels.
6054 Set spread usage of stereo image across Y axis for all channels.
6056 @item fcx, flx, frx, blx, brx, slx, srx, bcx
6057 Set spread usage of stereo image across X axis for each channel.
6059 @item fcy, fly, fry, bly, bry, sly, sry, bcy
6060 Set spread usage of stereo image across Y axis for each channel.
6063 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
6066 Set window function.
6068 It accepts the following values:
6091 Default is @code{hann}.
6094 Set window overlap. If set to 1, the recommended overlap for selected
6095 window function will be picked. Default is @code{0.5}.
6098 @section treble, highshelf
6100 Boost or cut treble (upper) frequencies of the audio using a two-pole
6101 shelving filter with a response similar to that of a standard
6102 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
6104 The filter accepts the following options:
6108 Give the gain at whichever is the lower of ~22 kHz and the
6109 Nyquist frequency. Its useful range is about -20 (for a large cut)
6110 to +20 (for a large boost). Beware of clipping when using a positive gain.
6113 Set the filter's central frequency and so can be used
6114 to extend or reduce the frequency range to be boosted or cut.
6115 The default value is @code{3000} Hz.
6118 Set method to specify band-width of filter.
6133 Determine how steep is the filter's shelf transition.
6136 Set number of poles. Default is 2.
6139 How much to use filtered signal in output. Default is 1.
6140 Range is between 0 and 1.
6143 Specify which channels to filter, by default all available are filtered.
6146 Normalize biquad coefficients, by default is disabled.
6147 Enabling it will normalize magnitude response at DC to 0dB.
6150 Set transform type of IIR filter.
6159 Set precison of filtering.
6162 Pick automatic sample format depending on surround filters.
6164 Always use signed 16-bit.
6166 Always use signed 32-bit.
6168 Always use float 32-bit.
6170 Always use float 64-bit.
6174 @subsection Commands
6176 This filter supports the following commands:
6179 Change treble frequency.
6180 Syntax for the command is : "@var{frequency}"
6183 Change treble width_type.
6184 Syntax for the command is : "@var{width_type}"
6187 Change treble width.
6188 Syntax for the command is : "@var{width}"
6192 Syntax for the command is : "@var{gain}"
6196 Syntax for the command is : "@var{mix}"
6201 Sinusoidal amplitude modulation.
6203 The filter accepts the following options:
6207 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
6208 (20 Hz or lower) will result in a tremolo effect.
6209 This filter may also be used as a ring modulator by specifying
6210 a modulation frequency higher than 20 Hz.
6211 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6214 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6215 Default value is 0.5.
6220 Sinusoidal phase modulation.
6222 The filter accepts the following options:
6226 Modulation frequency in Hertz.
6227 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6230 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6231 Default value is 0.5.
6236 Adjust the input audio volume.
6238 It accepts the following parameters:
6242 Set audio volume expression.
6244 Output values are clipped to the maximum value.
6246 The output audio volume is given by the relation:
6248 @var{output_volume} = @var{volume} * @var{input_volume}
6251 The default value for @var{volume} is "1.0".
6254 This parameter represents the mathematical precision.
6256 It determines which input sample formats will be allowed, which affects the
6257 precision of the volume scaling.
6261 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
6263 32-bit floating-point; this limits input sample format to FLT. (default)
6265 64-bit floating-point; this limits input sample format to DBL.
6269 Choose the behaviour on encountering ReplayGain side data in input frames.
6273 Remove ReplayGain side data, ignoring its contents (the default).
6276 Ignore ReplayGain side data, but leave it in the frame.
6279 Prefer the track gain, if present.
6282 Prefer the album gain, if present.
6285 @item replaygain_preamp
6286 Pre-amplification gain in dB to apply to the selected replaygain gain.
6288 Default value for @var{replaygain_preamp} is 0.0.
6290 @item replaygain_noclip
6291 Prevent clipping by limiting the gain applied.
6293 Default value for @var{replaygain_noclip} is 1.
6296 Set when the volume expression is evaluated.
6298 It accepts the following values:
6301 only evaluate expression once during the filter initialization, or
6302 when the @samp{volume} command is sent
6305 evaluate expression for each incoming frame
6308 Default value is @samp{once}.
6311 The volume expression can contain the following parameters.
6315 frame number (starting at zero)
6318 @item nb_consumed_samples
6319 number of samples consumed by the filter
6321 number of samples in the current frame
6323 original frame position in the file
6329 PTS at start of stream
6331 time at start of stream
6337 last set volume value
6340 Note that when @option{eval} is set to @samp{once} only the
6341 @var{sample_rate} and @var{tb} variables are available, all other
6342 variables will evaluate to NAN.
6344 @subsection Commands
6346 This filter supports the following commands:
6349 Modify the volume expression.
6350 The command accepts the same syntax of the corresponding option.
6352 If the specified expression is not valid, it is kept at its current
6356 @subsection Examples
6360 Halve the input audio volume:
6364 volume=volume=-6.0206dB
6367 In all the above example the named key for @option{volume} can be
6368 omitted, for example like in:
6374 Increase input audio power by 6 decibels using fixed-point precision:
6376 volume=volume=6dB:precision=fixed
6380 Fade volume after time 10 with an annihilation period of 5 seconds:
6382 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
6386 @section volumedetect
6388 Detect the volume of the input video.
6390 The filter has no parameters. The input is not modified. Statistics about
6391 the volume will be printed in the log when the input stream end is reached.
6393 In particular it will show the mean volume (root mean square), maximum
6394 volume (on a per-sample basis), and the beginning of a histogram of the
6395 registered volume values (from the maximum value to a cumulated 1/1000 of
6398 All volumes are in decibels relative to the maximum PCM value.
6400 @subsection Examples
6402 Here is an excerpt of the output:
6404 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
6405 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
6406 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
6407 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
6408 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
6409 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
6410 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
6411 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
6412 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
6418 The mean square energy is approximately -27 dB, or 10^-2.7.
6420 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
6422 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
6425 In other words, raising the volume by +4 dB does not cause any clipping,
6426 raising it by +5 dB causes clipping for 6 samples, etc.
6428 @c man end AUDIO FILTERS
6430 @chapter Audio Sources
6431 @c man begin AUDIO SOURCES
6433 Below is a description of the currently available audio sources.
6437 Buffer audio frames, and make them available to the filter chain.
6439 This source is mainly intended for a programmatic use, in particular
6440 through the interface defined in @file{libavfilter/buffersrc.h}.
6442 It accepts the following parameters:
6446 The timebase which will be used for timestamps of submitted frames. It must be
6447 either a floating-point number or in @var{numerator}/@var{denominator} form.
6450 The sample rate of the incoming audio buffers.
6453 The sample format of the incoming audio buffers.
6454 Either a sample format name or its corresponding integer representation from
6455 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
6457 @item channel_layout
6458 The channel layout of the incoming audio buffers.
6459 Either a channel layout name from channel_layout_map in
6460 @file{libavutil/channel_layout.c} or its corresponding integer representation
6461 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
6464 The number of channels of the incoming audio buffers.
6465 If both @var{channels} and @var{channel_layout} are specified, then they
6470 @subsection Examples
6473 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
6476 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
6477 Since the sample format with name "s16p" corresponds to the number
6478 6 and the "stereo" channel layout corresponds to the value 0x3, this is
6481 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
6486 Generate an audio signal specified by an expression.
6488 This source accepts in input one or more expressions (one for each
6489 channel), which are evaluated and used to generate a corresponding
6492 This source accepts the following options:
6496 Set the '|'-separated expressions list for each separate channel. In case the
6497 @option{channel_layout} option is not specified, the selected channel layout
6498 depends on the number of provided expressions. Otherwise the last
6499 specified expression is applied to the remaining output channels.
6501 @item channel_layout, c
6502 Set the channel layout. The number of channels in the specified layout
6503 must be equal to the number of specified expressions.
6506 Set the minimum duration of the sourced audio. See
6507 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6508 for the accepted syntax.
6509 Note that the resulting duration may be greater than the specified
6510 duration, as the generated audio is always cut at the end of a
6513 If not specified, or the expressed duration is negative, the audio is
6514 supposed to be generated forever.
6517 Set the number of samples per channel per each output frame,
6520 @item sample_rate, s
6521 Specify the sample rate, default to 44100.
6524 Each expression in @var{exprs} can contain the following constants:
6528 number of the evaluated sample, starting from 0
6531 time of the evaluated sample expressed in seconds, starting from 0
6538 @subsection Examples
6548 Generate a sin signal with frequency of 440 Hz, set sample rate to
6551 aevalsrc="sin(440*2*PI*t):s=8000"
6555 Generate a two channels signal, specify the channel layout (Front
6556 Center + Back Center) explicitly:
6558 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
6562 Generate white noise:
6564 aevalsrc="-2+random(0)"
6568 Generate an amplitude modulated signal:
6570 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
6574 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
6576 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
6583 Generate a FIR coefficients using frequency sampling method.
6585 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6587 The filter accepts the following options:
6591 Set number of filter coefficents in output audio stream.
6592 Default value is 1025.
6595 Set frequency points from where magnitude and phase are set.
6596 This must be in non decreasing order, and first element must be 0, while last element
6597 must be 1. Elements are separated by white spaces.
6600 Set magnitude value for every frequency point set by @option{frequency}.
6601 Number of values must be same as number of frequency points.
6602 Values are separated by white spaces.
6605 Set phase value for every frequency point set by @option{frequency}.
6606 Number of values must be same as number of frequency points.
6607 Values are separated by white spaces.
6609 @item sample_rate, r
6610 Set sample rate, default is 44100.
6613 Set number of samples per each frame. Default is 1024.
6616 Set window function. Default is blackman.
6621 The null audio source, return unprocessed audio frames. It is mainly useful
6622 as a template and to be employed in analysis / debugging tools, or as
6623 the source for filters which ignore the input data (for example the sox
6626 This source accepts the following options:
6630 @item channel_layout, cl
6632 Specifies the channel layout, and can be either an integer or a string
6633 representing a channel layout. The default value of @var{channel_layout}
6636 Check the channel_layout_map definition in
6637 @file{libavutil/channel_layout.c} for the mapping between strings and
6638 channel layout values.
6640 @item sample_rate, r
6641 Specifies the sample rate, and defaults to 44100.
6644 Set the number of samples per requested frames.
6647 Set the duration of the sourced audio. See
6648 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6649 for the accepted syntax.
6651 If not specified, or the expressed duration is negative, the audio is
6652 supposed to be generated forever.
6655 @subsection Examples
6659 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
6661 anullsrc=r=48000:cl=4
6665 Do the same operation with a more obvious syntax:
6667 anullsrc=r=48000:cl=mono
6671 All the parameters need to be explicitly defined.
6675 Synthesize a voice utterance using the libflite library.
6677 To enable compilation of this filter you need to configure FFmpeg with
6678 @code{--enable-libflite}.
6680 Note that versions of the flite library prior to 2.0 are not thread-safe.
6682 The filter accepts the following options:
6687 If set to 1, list the names of the available voices and exit
6688 immediately. Default value is 0.
6691 Set the maximum number of samples per frame. Default value is 512.
6694 Set the filename containing the text to speak.
6697 Set the text to speak.
6700 Set the voice to use for the speech synthesis. Default value is
6701 @code{kal}. See also the @var{list_voices} option.
6704 @subsection Examples
6708 Read from file @file{speech.txt}, and synthesize the text using the
6709 standard flite voice:
6711 flite=textfile=speech.txt
6715 Read the specified text selecting the @code{slt} voice:
6717 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6721 Input text to ffmpeg:
6723 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6727 Make @file{ffplay} speak the specified text, using @code{flite} and
6728 the @code{lavfi} device:
6730 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
6734 For more information about libflite, check:
6735 @url{http://www.festvox.org/flite/}
6739 Generate a noise audio signal.
6741 The filter accepts the following options:
6744 @item sample_rate, r
6745 Specify the sample rate. Default value is 48000 Hz.
6748 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
6752 Specify the duration of the generated audio stream. Not specifying this option
6753 results in noise with an infinite length.
6755 @item color, colour, c
6756 Specify the color of noise. Available noise colors are white, pink, brown,
6757 blue, violet and velvet. Default color is white.
6760 Specify a value used to seed the PRNG.
6763 Set the number of samples per each output frame, default is 1024.
6766 @subsection Examples
6771 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
6773 anoisesrc=d=60:c=pink:r=44100:a=0.5
6779 Generate odd-tap Hilbert transform FIR coefficients.
6781 The resulting stream can be used with @ref{afir} filter for phase-shifting
6782 the signal by 90 degrees.
6784 This is used in many matrix coding schemes and for analytic signal generation.
6785 The process is often written as a multiplication by i (or j), the imaginary unit.
6787 The filter accepts the following options:
6791 @item sample_rate, s
6792 Set sample rate, default is 44100.
6795 Set length of FIR filter, default is 22051.
6798 Set number of samples per each frame.
6801 Set window function to be used when generating FIR coefficients.
6806 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
6808 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6810 The filter accepts the following options:
6813 @item sample_rate, r
6814 Set sample rate, default is 44100.
6817 Set number of samples per each frame. Default is 1024.
6820 Set high-pass frequency. Default is 0.
6823 Set low-pass frequency. Default is 0.
6824 If high-pass frequency is lower than low-pass frequency and low-pass frequency
6825 is higher than 0 then filter will create band-pass filter coefficients,
6826 otherwise band-reject filter coefficients.
6829 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
6832 Set Kaiser window beta.
6835 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
6838 Enable rounding, by default is disabled.
6841 Set number of taps for high-pass filter.
6844 Set number of taps for low-pass filter.
6849 Generate an audio signal made of a sine wave with amplitude 1/8.
6851 The audio signal is bit-exact.
6853 The filter accepts the following options:
6858 Set the carrier frequency. Default is 440 Hz.
6860 @item beep_factor, b
6861 Enable a periodic beep every second with frequency @var{beep_factor} times
6862 the carrier frequency. Default is 0, meaning the beep is disabled.
6864 @item sample_rate, r
6865 Specify the sample rate, default is 44100.
6868 Specify the duration of the generated audio stream.
6870 @item samples_per_frame
6871 Set the number of samples per output frame.
6873 The expression can contain the following constants:
6877 The (sequential) number of the output audio frame, starting from 0.
6880 The PTS (Presentation TimeStamp) of the output audio frame,
6881 expressed in @var{TB} units.
6884 The PTS of the output audio frame, expressed in seconds.
6887 The timebase of the output audio frames.
6890 Default is @code{1024}.
6893 @subsection Examples
6898 Generate a simple 440 Hz sine wave:
6904 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
6908 sine=frequency=220:beep_factor=4:duration=5
6912 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
6915 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
6919 @c man end AUDIO SOURCES
6921 @chapter Audio Sinks
6922 @c man begin AUDIO SINKS
6924 Below is a description of the currently available audio sinks.
6926 @section abuffersink
6928 Buffer audio frames, and make them available to the end of filter chain.
6930 This sink is mainly intended for programmatic use, in particular
6931 through the interface defined in @file{libavfilter/buffersink.h}
6932 or the options system.
6934 It accepts a pointer to an AVABufferSinkContext structure, which
6935 defines the incoming buffers' formats, to be passed as the opaque
6936 parameter to @code{avfilter_init_filter} for initialization.
6939 Null audio sink; do absolutely nothing with the input audio. It is
6940 mainly useful as a template and for use in analysis / debugging
6943 @c man end AUDIO SINKS
6945 @chapter Video Filters
6946 @c man begin VIDEO FILTERS
6948 When you configure your FFmpeg build, you can disable any of the
6949 existing filters using @code{--disable-filters}.
6950 The configure output will show the video filters included in your
6953 Below is a description of the currently available video filters.
6957 Mark a region of interest in a video frame.
6959 The frame data is passed through unchanged, but metadata is attached
6960 to the frame indicating regions of interest which can affect the
6961 behaviour of later encoding. Multiple regions can be marked by
6962 applying the filter multiple times.
6966 Region distance in pixels from the left edge of the frame.
6968 Region distance in pixels from the top edge of the frame.
6970 Region width in pixels.
6972 Region height in pixels.
6974 The parameters @var{x}, @var{y}, @var{w} and @var{h} are expressions,
6975 and may contain the following variables:
6978 Width of the input frame.
6980 Height of the input frame.
6984 Quantisation offset to apply within the region.
6986 This must be a real value in the range -1 to +1. A value of zero
6987 indicates no quality change. A negative value asks for better quality
6988 (less quantisation), while a positive value asks for worse quality
6989 (greater quantisation).
6991 The range is calibrated so that the extreme values indicate the
6992 largest possible offset - if the rest of the frame is encoded with the
6993 worst possible quality, an offset of -1 indicates that this region
6994 should be encoded with the best possible quality anyway. Intermediate
6995 values are then interpolated in some codec-dependent way.
6997 For example, in 10-bit H.264 the quantisation parameter varies between
6998 -12 and 51. A typical qoffset value of -1/10 therefore indicates that
6999 this region should be encoded with a QP around one-tenth of the full
7000 range better than the rest of the frame. So, if most of the frame
7001 were to be encoded with a QP of around 30, this region would get a QP
7002 of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3).
7003 An extreme value of -1 would indicate that this region should be
7004 encoded with the best possible quality regardless of the treatment of
7005 the rest of the frame - that is, should be encoded at a QP of -12.
7007 If set to true, remove any existing regions of interest marked on the
7008 frame before adding the new one.
7011 @subsection Examples
7015 Mark the centre quarter of the frame as interesting.
7017 addroi=iw/4:ih/4:iw/2:ih/2:-1/10
7020 Mark the 100-pixel-wide region on the left edge of the frame as very
7021 uninteresting (to be encoded at much lower quality than the rest of
7024 addroi=0:0:100:ih:+1/5
7028 @section alphaextract
7030 Extract the alpha component from the input as a grayscale video. This
7031 is especially useful with the @var{alphamerge} filter.
7035 Add or replace the alpha component of the primary input with the
7036 grayscale value of a second input. This is intended for use with
7037 @var{alphaextract} to allow the transmission or storage of frame
7038 sequences that have alpha in a format that doesn't support an alpha
7041 For example, to reconstruct full frames from a normal YUV-encoded video
7042 and a separate video created with @var{alphaextract}, you might use:
7044 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
7049 Amplify differences between current pixel and pixels of adjacent frames in
7050 same pixel location.
7052 This filter accepts the following options:
7056 Set frame radius. Default is 2. Allowed range is from 1 to 63.
7057 For example radius of 3 will instruct filter to calculate average of 7 frames.
7060 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
7063 Set threshold for difference amplification. Any difference greater or equal to
7064 this value will not alter source pixel. Default is 10.
7065 Allowed range is from 0 to 65535.
7068 Set tolerance for difference amplification. Any difference lower to
7069 this value will not alter source pixel. Default is 0.
7070 Allowed range is from 0 to 65535.
7073 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7074 This option controls maximum possible value that will decrease source pixel value.
7077 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7078 This option controls maximum possible value that will increase source pixel value.
7081 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
7084 @subsection Commands
7086 This filter supports the following @ref{commands} that corresponds to option of same name:
7098 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
7099 and libavformat to work. On the other hand, it is limited to ASS (Advanced
7100 Substation Alpha) subtitles files.
7102 This filter accepts the following option in addition to the common options from
7103 the @ref{subtitles} filter:
7107 Set the shaping engine
7109 Available values are:
7112 The default libass shaping engine, which is the best available.
7114 Fast, font-agnostic shaper that can do only substitutions
7116 Slower shaper using OpenType for substitutions and positioning
7119 The default is @code{auto}.
7123 Apply an Adaptive Temporal Averaging Denoiser to the video input.
7125 The filter accepts the following options:
7129 Set threshold A for 1st plane. Default is 0.02.
7130 Valid range is 0 to 0.3.
7133 Set threshold B for 1st plane. Default is 0.04.
7134 Valid range is 0 to 5.
7137 Set threshold A for 2nd plane. Default is 0.02.
7138 Valid range is 0 to 0.3.
7141 Set threshold B for 2nd plane. Default is 0.04.
7142 Valid range is 0 to 5.
7145 Set threshold A for 3rd plane. Default is 0.02.
7146 Valid range is 0 to 0.3.
7149 Set threshold B for 3rd plane. Default is 0.04.
7150 Valid range is 0 to 5.
7152 Threshold A is designed to react on abrupt changes in the input signal and
7153 threshold B is designed to react on continuous changes in the input signal.
7156 Set number of frames filter will use for averaging. Default is 9. Must be odd
7157 number in range [5, 129].
7160 Set what planes of frame filter will use for averaging. Default is all.
7163 Set what variant of algorithm filter will use for averaging. Default is @code{p} parallel.
7164 Alternatively can be set to @code{s} serial.
7166 Parallel can be faster then serial, while other way around is never true.
7167 Parallel will abort early on first change being greater then thresholds, while serial
7168 will continue processing other side of frames if they are equal or below thresholds.
7173 Set sigma for 1st plane, 2nd plane or 3rd plane. Default is 32767.
7174 Valid range is from 0 to 32767.
7175 This options controls weight for each pixel in radius defined by size.
7176 Default value means every pixel have same weight.
7177 Setting this option to 0 effectively disables filtering.
7180 @subsection Commands
7181 This filter supports same @ref{commands} as options except option @code{s}.
7182 The command accepts the same syntax of the corresponding option.
7186 Apply average blur filter.
7188 The filter accepts the following options:
7192 Set horizontal radius size.
7195 Set which planes to filter. By default all planes are filtered.
7198 Set vertical radius size, if zero it will be same as @code{sizeX}.
7199 Default is @code{0}.
7202 @subsection Commands
7203 This filter supports same commands as options.
7204 The command accepts the same syntax of the corresponding option.
7206 If the specified expression is not valid, it is kept at its current
7211 Compute the bounding box for the non-black pixels in the input frame
7214 This filter computes the bounding box containing all the pixels with a
7215 luminance value greater than the minimum allowed value.
7216 The parameters describing the bounding box are printed on the filter
7219 The filter accepts the following option:
7223 Set the minimal luminance value. Default is @code{16}.
7226 @subsection Commands
7228 This filter supports the all above options as @ref{commands}.
7231 Apply bilateral filter, spatial smoothing while preserving edges.
7233 The filter accepts the following options:
7236 Set sigma of gaussian function to calculate spatial weight.
7237 Allowed range is 0 to 512. Default is 0.1.
7240 Set sigma of gaussian function to calculate range weight.
7241 Allowed range is 0 to 1. Default is 0.1.
7244 Set planes to filter. Default is first only.
7247 @subsection Commands
7249 This filter supports the all above options as @ref{commands}.
7251 @section bitplanenoise
7253 Show and measure bit plane noise.
7255 The filter accepts the following options:
7259 Set which plane to analyze. Default is @code{1}.
7262 Filter out noisy pixels from @code{bitplane} set above.
7263 Default is disabled.
7266 @section blackdetect
7268 Detect video intervals that are (almost) completely black. Can be
7269 useful to detect chapter transitions, commercials, or invalid
7272 The filter outputs its detection analysis to both the log as well as
7273 frame metadata. If a black segment of at least the specified minimum
7274 duration is found, a line with the start and end timestamps as well
7275 as duration is printed to the log with level @code{info}. In addition,
7276 a log line with level @code{debug} is printed per frame showing the
7277 black amount detected for that frame.
7279 The filter also attaches metadata to the first frame of a black
7280 segment with key @code{lavfi.black_start} and to the first frame
7281 after the black segment ends with key @code{lavfi.black_end}. The
7282 value is the frame's timestamp. This metadata is added regardless
7283 of the minimum duration specified.
7285 The filter accepts the following options:
7288 @item black_min_duration, d
7289 Set the minimum detected black duration expressed in seconds. It must
7290 be a non-negative floating point number.
7292 Default value is 2.0.
7294 @item picture_black_ratio_th, pic_th
7295 Set the threshold for considering a picture "black".
7296 Express the minimum value for the ratio:
7298 @var{nb_black_pixels} / @var{nb_pixels}
7301 for which a picture is considered black.
7302 Default value is 0.98.
7304 @item pixel_black_th, pix_th
7305 Set the threshold for considering a pixel "black".
7307 The threshold expresses the maximum pixel luminance value for which a
7308 pixel is considered "black". The provided value is scaled according to
7309 the following equation:
7311 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
7314 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
7315 the input video format, the range is [0-255] for YUV full-range
7316 formats and [16-235] for YUV non full-range formats.
7318 Default value is 0.10.
7321 The following example sets the maximum pixel threshold to the minimum
7322 value, and detects only black intervals of 2 or more seconds:
7324 blackdetect=d=2:pix_th=0.00
7329 Detect frames that are (almost) completely black. Can be useful to
7330 detect chapter transitions or commercials. Output lines consist of
7331 the frame number of the detected frame, the percentage of blackness,
7332 the position in the file if known or -1 and the timestamp in seconds.
7334 In order to display the output lines, you need to set the loglevel at
7335 least to the AV_LOG_INFO value.
7337 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
7338 The value represents the percentage of pixels in the picture that
7339 are below the threshold value.
7341 It accepts the following parameters:
7346 The percentage of the pixels that have to be below the threshold; it defaults to
7349 @item threshold, thresh
7350 The threshold below which a pixel value is considered black; it defaults to
7358 Blend two video frames into each other.
7360 The @code{blend} filter takes two input streams and outputs one
7361 stream, the first input is the "top" layer and second input is
7362 "bottom" layer. By default, the output terminates when the longest input terminates.
7364 The @code{tblend} (time blend) filter takes two consecutive frames
7365 from one single stream, and outputs the result obtained by blending
7366 the new frame on top of the old frame.
7368 A description of the accepted options follows.
7376 Set blend mode for specific pixel component or all pixel components in case
7377 of @var{all_mode}. Default value is @code{normal}.
7379 Available values for component modes are:
7421 Set blend opacity for specific pixel component or all pixel components in case
7422 of @var{all_opacity}. Only used in combination with pixel component blend modes.
7429 Set blend expression for specific pixel component or all pixel components in case
7430 of @var{all_expr}. Note that related mode options will be ignored if those are set.
7432 The expressions can use the following variables:
7436 The sequential number of the filtered frame, starting from @code{0}.
7440 the coordinates of the current sample
7444 the width and height of currently filtered plane
7448 Width and height scale for the plane being filtered. It is the
7449 ratio between the dimensions of the current plane to the luma plane,
7450 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
7451 the luma plane and @code{0.5,0.5} for the chroma planes.
7454 Time of the current frame, expressed in seconds.
7457 Value of pixel component at current location for first video frame (top layer).
7460 Value of pixel component at current location for second video frame (bottom layer).
7464 The @code{blend} filter also supports the @ref{framesync} options.
7466 @subsection Examples
7470 Apply transition from bottom layer to top layer in first 10 seconds:
7472 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
7476 Apply linear horizontal transition from top layer to bottom layer:
7478 blend=all_expr='A*(X/W)+B*(1-X/W)'
7482 Apply 1x1 checkerboard effect:
7484 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
7488 Apply uncover left effect:
7490 blend=all_expr='if(gte(N*SW+X,W),A,B)'
7494 Apply uncover down effect:
7496 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
7500 Apply uncover up-left effect:
7502 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
7506 Split diagonally video and shows top and bottom layer on each side:
7508 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
7512 Display differences between the current and the previous frame:
7514 tblend=all_mode=grainextract
7518 @subsection Commands
7519 This filter supports same @ref{commands} as options.
7523 Denoise frames using Block-Matching 3D algorithm.
7525 The filter accepts the following options.
7529 Set denoising strength. Default value is 1.
7530 Allowed range is from 0 to 999.9.
7531 The denoising algorithm is very sensitive to sigma, so adjust it
7532 according to the source.
7535 Set local patch size. This sets dimensions in 2D.
7538 Set sliding step for processing blocks. Default value is 4.
7539 Allowed range is from 1 to 64.
7540 Smaller values allows processing more reference blocks and is slower.
7543 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
7544 When set to 1, no block matching is done. Larger values allows more blocks
7546 Allowed range is from 1 to 256.
7549 Set radius for search block matching. Default is 9.
7550 Allowed range is from 1 to INT32_MAX.
7553 Set step between two search locations for block matching. Default is 1.
7554 Allowed range is from 1 to 64. Smaller is slower.
7557 Set threshold of mean square error for block matching. Valid range is 0 to
7561 Set thresholding parameter for hard thresholding in 3D transformed domain.
7562 Larger values results in stronger hard-thresholding filtering in frequency
7566 Set filtering estimation mode. Can be @code{basic} or @code{final}.
7567 Default is @code{basic}.
7570 If enabled, filter will use 2nd stream for block matching.
7571 Default is disabled for @code{basic} value of @var{estim} option,
7572 and always enabled if value of @var{estim} is @code{final}.
7575 Set planes to filter. Default is all available except alpha.
7578 @subsection Examples
7582 Basic filtering with bm3d:
7584 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
7588 Same as above, but filtering only luma:
7590 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
7594 Same as above, but with both estimation modes:
7596 split[a][b],[a]bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
7600 Same as above, but prefilter with @ref{nlmeans} filter instead:
7602 split[a][b],[a]nlmeans=s=3:r=7:p=3[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
7608 Apply a boxblur algorithm to the input video.
7610 It accepts the following parameters:
7614 @item luma_radius, lr
7615 @item luma_power, lp
7616 @item chroma_radius, cr
7617 @item chroma_power, cp
7618 @item alpha_radius, ar
7619 @item alpha_power, ap
7623 A description of the accepted options follows.
7626 @item luma_radius, lr
7627 @item chroma_radius, cr
7628 @item alpha_radius, ar
7629 Set an expression for the box radius in pixels used for blurring the
7630 corresponding input plane.
7632 The radius value must be a non-negative number, and must not be
7633 greater than the value of the expression @code{min(w,h)/2} for the
7634 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
7637 Default value for @option{luma_radius} is "2". If not specified,
7638 @option{chroma_radius} and @option{alpha_radius} default to the
7639 corresponding value set for @option{luma_radius}.
7641 The expressions can contain the following constants:
7645 The input width and height in pixels.
7649 The input chroma image width and height in pixels.
7653 The horizontal and vertical chroma subsample values. For example, for the
7654 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
7657 @item luma_power, lp
7658 @item chroma_power, cp
7659 @item alpha_power, ap
7660 Specify how many times the boxblur filter is applied to the
7661 corresponding plane.
7663 Default value for @option{luma_power} is 2. If not specified,
7664 @option{chroma_power} and @option{alpha_power} default to the
7665 corresponding value set for @option{luma_power}.
7667 A value of 0 will disable the effect.
7670 @subsection Examples
7674 Apply a boxblur filter with the luma, chroma, and alpha radii
7677 boxblur=luma_radius=2:luma_power=1
7682 Set the luma radius to 2, and alpha and chroma radius to 0:
7684 boxblur=2:1:cr=0:ar=0
7688 Set the luma and chroma radii to a fraction of the video dimension:
7690 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
7696 Deinterlace the input video ("bwdif" stands for "Bob Weaver
7697 Deinterlacing Filter").
7699 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
7700 interpolation algorithms.
7701 It accepts the following parameters:
7705 The interlacing mode to adopt. It accepts one of the following values:
7709 Output one frame for each frame.
7711 Output one frame for each field.
7714 The default value is @code{send_field}.
7717 The picture field parity assumed for the input interlaced video. It accepts one
7718 of the following values:
7722 Assume the top field is first.
7724 Assume the bottom field is first.
7726 Enable automatic detection of field parity.
7729 The default value is @code{auto}.
7730 If the interlacing is unknown or the decoder does not export this information,
7731 top field first will be assumed.
7734 Specify which frames to deinterlace. Accepts one of the following
7739 Deinterlace all frames.
7741 Only deinterlace frames marked as interlaced.
7744 The default value is @code{all}.
7749 Apply Contrast Adaptive Sharpen filter to video stream.
7751 The filter accepts the following options:
7755 Set the sharpening strength. Default value is 0.
7758 Set planes to filter. Default value is to filter all
7759 planes except alpha plane.
7762 @subsection Commands
7763 This filter supports same @ref{commands} as options.
7766 Remove all color information for all colors except for certain one.
7768 The filter accepts the following options:
7772 The color which will not be replaced with neutral chroma.
7775 Similarity percentage with the above color.
7776 0.01 matches only the exact key color, while 1.0 matches everything.
7780 0.0 makes pixels either fully gray, or not gray at all.
7781 Higher values result in more preserved color.
7784 Signals that the color passed is already in YUV instead of RGB.
7786 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7787 This can be used to pass exact YUV values as hexadecimal numbers.
7790 @subsection Commands
7791 This filter supports same @ref{commands} as options.
7792 The command accepts the same syntax of the corresponding option.
7794 If the specified expression is not valid, it is kept at its current
7798 YUV colorspace color/chroma keying.
7800 The filter accepts the following options:
7804 The color which will be replaced with transparency.
7807 Similarity percentage with the key color.
7809 0.01 matches only the exact key color, while 1.0 matches everything.
7814 0.0 makes pixels either fully transparent, or not transparent at all.
7816 Higher values result in semi-transparent pixels, with a higher transparency
7817 the more similar the pixels color is to the key color.
7820 Signals that the color passed is already in YUV instead of RGB.
7822 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7823 This can be used to pass exact YUV values as hexadecimal numbers.
7826 @subsection Commands
7827 This filter supports same @ref{commands} as options.
7828 The command accepts the same syntax of the corresponding option.
7830 If the specified expression is not valid, it is kept at its current
7833 @subsection Examples
7837 Make every green pixel in the input image transparent:
7839 ffmpeg -i input.png -vf chromakey=green out.png
7843 Overlay a greenscreen-video on top of a static black background.
7845 ffmpeg -f lavfi -i color=c=black:s=1280x720 -i video.mp4 -shortest -filter_complex "[1:v]chromakey=0x70de77:0.1:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.mkv
7850 Reduce chrominance noise.
7852 The filter accepts the following options:
7856 Set threshold for averaging chrominance values.
7857 Sum of absolute difference of Y, U and V pixel components of current
7858 pixel and neighbour pixels lower than this threshold will be used in
7859 averaging. Luma component is left unchanged and is copied to output.
7860 Default value is 30. Allowed range is from 1 to 200.
7863 Set horizontal radius of rectangle used for averaging.
7864 Allowed range is from 1 to 100. Default value is 5.
7867 Set vertical radius of rectangle used for averaging.
7868 Allowed range is from 1 to 100. Default value is 5.
7871 Set horizontal step when averaging. Default value is 1.
7872 Allowed range is from 1 to 50.
7873 Mostly useful to speed-up filtering.
7876 Set vertical step when averaging. Default value is 1.
7877 Allowed range is from 1 to 50.
7878 Mostly useful to speed-up filtering.
7881 Set Y threshold for averaging chrominance values.
7882 Set finer control for max allowed difference between Y components
7883 of current pixel and neigbour pixels.
7884 Default value is 200. Allowed range is from 1 to 200.
7887 Set U threshold for averaging chrominance values.
7888 Set finer control for max allowed difference between U components
7889 of current pixel and neigbour pixels.
7890 Default value is 200. Allowed range is from 1 to 200.
7893 Set V threshold for averaging chrominance values.
7894 Set finer control for max allowed difference between V components
7895 of current pixel and neigbour pixels.
7896 Default value is 200. Allowed range is from 1 to 200.
7899 @subsection Commands
7900 This filter supports same @ref{commands} as options.
7901 The command accepts the same syntax of the corresponding option.
7903 @section chromashift
7904 Shift chroma pixels horizontally and/or vertically.
7906 The filter accepts the following options:
7909 Set amount to shift chroma-blue horizontally.
7911 Set amount to shift chroma-blue vertically.
7913 Set amount to shift chroma-red horizontally.
7915 Set amount to shift chroma-red vertically.
7917 Set edge mode, can be @var{smear}, default, or @var{warp}.
7920 @subsection Commands
7922 This filter supports the all above options as @ref{commands}.
7926 Display CIE color diagram with pixels overlaid onto it.
7928 The filter accepts the following options:
7943 @item uhdtv, rec2020
7957 Set what gamuts to draw.
7959 See @code{system} option for available values.
7962 Set ciescope size, by default set to 512.
7965 Set intensity used to map input pixel values to CIE diagram.
7968 Set contrast used to draw tongue colors that are out of active color system gamut.
7971 Correct gamma displayed on scope, by default enabled.
7974 Show white point on CIE diagram, by default disabled.
7977 Set input gamma. Used only with XYZ input color space.
7982 Visualize information exported by some codecs.
7984 Some codecs can export information through frames using side-data or other
7985 means. For example, some MPEG based codecs export motion vectors through the
7986 @var{export_mvs} flag in the codec @option{flags2} option.
7988 The filter accepts the following option:
7992 Set motion vectors to visualize.
7994 Available flags for @var{mv} are:
7998 forward predicted MVs of P-frames
8000 forward predicted MVs of B-frames
8002 backward predicted MVs of B-frames
8006 Display quantization parameters using the chroma planes.
8009 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
8011 Available flags for @var{mv_type} are:
8015 forward predicted MVs
8017 backward predicted MVs
8020 @item frame_type, ft
8021 Set frame type to visualize motion vectors of.
8023 Available flags for @var{frame_type} are:
8027 intra-coded frames (I-frames)
8029 predicted frames (P-frames)
8031 bi-directionally predicted frames (B-frames)
8035 @subsection Examples
8039 Visualize forward predicted MVs of all frames using @command{ffplay}:
8041 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
8045 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
8047 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
8051 @section colorbalance
8052 Modify intensity of primary colors (red, green and blue) of input frames.
8054 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
8055 regions for the red-cyan, green-magenta or blue-yellow balance.
8057 A positive adjustment value shifts the balance towards the primary color, a negative
8058 value towards the complementary color.
8060 The filter accepts the following options:
8066 Adjust red, green and blue shadows (darkest pixels).
8071 Adjust red, green and blue midtones (medium pixels).
8076 Adjust red, green and blue highlights (brightest pixels).
8078 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8081 Preserve lightness when changing color balance. Default is disabled.
8084 @subsection Examples
8088 Add red color cast to shadows:
8094 @subsection Commands
8096 This filter supports the all above options as @ref{commands}.
8098 @section colorcontrast
8100 Adjust color contrast between RGB components.
8102 The filter accepts the following options:
8106 Set the red-cyan contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8109 Set the green-magenta contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8112 Set the blue-yellow contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8117 Set the weight of each @code{rc}, @code{gm}, @code{by} option value. Default value is 0.0.
8118 Allowed range is from 0.0 to 1.0. If all weights are 0.0 filtering is disabled.
8121 Set the amount of preserving lightness. Default value is 0.0. Allowed range is from 0.0 to 1.0.
8124 @subsection Commands
8126 This filter supports the all above options as @ref{commands}.
8128 @section colorcorrect
8130 Adjust color white balance selectively for blacks and whites.
8131 This filter operates in YUV colorspace.
8133 The filter accepts the following options:
8137 Set the red shadow spot. Allowed range is from -1.0 to 1.0.
8141 Set the blue shadow spot. Allowed range is from -1.0 to 1.0.
8145 Set the red highlight spot. Allowed range is from -1.0 to 1.0.
8149 Set the red highlight spot. Allowed range is from -1.0 to 1.0.
8153 Set the amount of saturation. Allowed range is from -3.0 to 3.0.
8157 @subsection Commands
8159 This filter supports the all above options as @ref{commands}.
8161 @section colorchannelmixer
8163 Adjust video input frames by re-mixing color channels.
8165 This filter modifies a color channel by adding the values associated to
8166 the other channels of the same pixels. For example if the value to
8167 modify is red, the output value will be:
8169 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
8172 The filter accepts the following options:
8179 Adjust contribution of input red, green, blue and alpha channels for output red channel.
8180 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
8186 Adjust contribution of input red, green, blue and alpha channels for output green channel.
8187 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
8193 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
8194 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
8200 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
8201 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
8203 Allowed ranges for options are @code{[-2.0, 2.0]}.
8206 Preserve lightness when changing colors. Allowed range is from @code{[0.0, 1.0]}.
8207 Default is @code{0.0}, thus disabled.
8210 @subsection Examples
8214 Convert source to grayscale:
8216 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
8219 Simulate sepia tones:
8221 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
8225 @subsection Commands
8227 This filter supports the all above options as @ref{commands}.
8230 Overlay a solid color on the video stream.
8232 The filter accepts the following options:
8236 Set the color hue. Allowed range is from 0 to 360.
8240 Set the color saturation. Allowed range is from 0 to 1.
8241 Default value is 0.5.
8244 Set the color lightness. Allowed range is from 0 to 1.
8245 Default value is 0.5.
8248 Set the mix of source lightness. By default is set to 1.0.
8249 Allowed range is from 0.0 to 1.0.
8252 @subsection Commands
8254 This filter supports the all above options as @ref{commands}.
8257 RGB colorspace color keying.
8259 The filter accepts the following options:
8263 The color which will be replaced with transparency.
8266 Similarity percentage with the key color.
8268 0.01 matches only the exact key color, while 1.0 matches everything.
8273 0.0 makes pixels either fully transparent, or not transparent at all.
8275 Higher values result in semi-transparent pixels, with a higher transparency
8276 the more similar the pixels color is to the key color.
8279 @subsection Examples
8283 Make every green pixel in the input image transparent:
8285 ffmpeg -i input.png -vf colorkey=green out.png
8289 Overlay a greenscreen-video on top of a static background image.
8291 ffmpeg -i background.png -i video.mp4 -filter_complex "[1:v]colorkey=0x3BBD1E:0.3:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.flv
8295 @subsection Commands
8296 This filter supports same @ref{commands} as options.
8297 The command accepts the same syntax of the corresponding option.
8299 If the specified expression is not valid, it is kept at its current
8303 Remove all color information for all RGB colors except for certain one.
8305 The filter accepts the following options:
8309 The color which will not be replaced with neutral gray.
8312 Similarity percentage with the above color.
8313 0.01 matches only the exact key color, while 1.0 matches everything.
8316 Blend percentage. 0.0 makes pixels fully gray.
8317 Higher values result in more preserved color.
8320 @subsection Commands
8321 This filter supports same @ref{commands} as options.
8322 The command accepts the same syntax of the corresponding option.
8324 If the specified expression is not valid, it is kept at its current
8327 @section colorlevels
8329 Adjust video input frames using levels.
8331 The filter accepts the following options:
8338 Adjust red, green, blue and alpha input black point.
8339 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8345 Adjust red, green, blue and alpha input white point.
8346 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
8348 Input levels are used to lighten highlights (bright tones), darken shadows
8349 (dark tones), change the balance of bright and dark tones.
8355 Adjust red, green, blue and alpha output black point.
8356 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
8362 Adjust red, green, blue and alpha output white point.
8363 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
8365 Output levels allows manual selection of a constrained output level range.
8368 @subsection Examples
8372 Make video output darker:
8374 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
8380 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
8384 Make video output lighter:
8386 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
8390 Increase brightness:
8392 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
8396 @subsection Commands
8398 This filter supports the all above options as @ref{commands}.
8400 @section colormatrix
8402 Convert color matrix.
8404 The filter accepts the following options:
8409 Specify the source and destination color matrix. Both values must be
8412 The accepted values are:
8440 For example to convert from BT.601 to SMPTE-240M, use the command:
8442 colormatrix=bt601:smpte240m
8447 Convert colorspace, transfer characteristics or color primaries.
8448 Input video needs to have an even size.
8450 The filter accepts the following options:
8455 Specify all color properties at once.
8457 The accepted values are:
8487 Specify output colorspace.
8489 The accepted values are:
8498 BT.470BG or BT.601-6 625
8501 SMPTE-170M or BT.601-6 525
8510 BT.2020 with non-constant luminance
8516 Specify output transfer characteristics.
8518 The accepted values are:
8530 Constant gamma of 2.2
8533 Constant gamma of 2.8
8536 SMPTE-170M, BT.601-6 625 or BT.601-6 525
8554 BT.2020 for 10-bits content
8557 BT.2020 for 12-bits content
8563 Specify output color primaries.
8565 The accepted values are:
8574 BT.470BG or BT.601-6 625
8577 SMPTE-170M or BT.601-6 525
8601 Specify output color range.
8603 The accepted values are:
8606 TV (restricted) range
8609 MPEG (restricted) range
8620 Specify output color format.
8622 The accepted values are:
8625 YUV 4:2:0 planar 8-bits
8628 YUV 4:2:0 planar 10-bits
8631 YUV 4:2:0 planar 12-bits
8634 YUV 4:2:2 planar 8-bits
8637 YUV 4:2:2 planar 10-bits
8640 YUV 4:2:2 planar 12-bits
8643 YUV 4:4:4 planar 8-bits
8646 YUV 4:4:4 planar 10-bits
8649 YUV 4:4:4 planar 12-bits
8654 Do a fast conversion, which skips gamma/primary correction. This will take
8655 significantly less CPU, but will be mathematically incorrect. To get output
8656 compatible with that produced by the colormatrix filter, use fast=1.
8659 Specify dithering mode.
8661 The accepted values are:
8667 Floyd-Steinberg dithering
8671 Whitepoint adaptation mode.
8673 The accepted values are:
8676 Bradford whitepoint adaptation
8679 von Kries whitepoint adaptation
8682 identity whitepoint adaptation (i.e. no whitepoint adaptation)
8686 Override all input properties at once. Same accepted values as @ref{all}.
8689 Override input colorspace. Same accepted values as @ref{space}.
8692 Override input color primaries. Same accepted values as @ref{primaries}.
8695 Override input transfer characteristics. Same accepted values as @ref{trc}.
8698 Override input color range. Same accepted values as @ref{range}.
8702 The filter converts the transfer characteristics, color space and color
8703 primaries to the specified user values. The output value, if not specified,
8704 is set to a default value based on the "all" property. If that property is
8705 also not specified, the filter will log an error. The output color range and
8706 format default to the same value as the input color range and format. The
8707 input transfer characteristics, color space, color primaries and color range
8708 should be set on the input data. If any of these are missing, the filter will
8709 log an error and no conversion will take place.
8711 For example to convert the input to SMPTE-240M, use the command:
8713 colorspace=smpte240m
8716 @section colortemperature
8717 Adjust color temperature in video to simulate variations in ambient color temperature.
8719 The filter accepts the following options:
8723 Set the temperature in Kelvin. Allowed range is from 1000 to 40000.
8724 Default value is 6500 K.
8727 Set mixing with filtered output. Allowed range is from 0 to 1.
8731 Set the amount of preserving lightness. Allowed range is from 0 to 1.
8735 @subsection Commands
8736 This filter supports same @ref{commands} as options.
8738 @section convolution
8740 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
8742 The filter accepts the following options:
8749 Set matrix for each plane.
8750 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
8751 and from 1 to 49 odd number of signed integers in @var{row} mode.
8757 Set multiplier for calculated value for each plane.
8758 If unset or 0, it will be sum of all matrix elements.
8764 Set bias for each plane. This value is added to the result of the multiplication.
8765 Useful for making the overall image brighter or darker. Default is 0.0.
8771 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
8772 Default is @var{square}.
8775 @subsection Commands
8777 This filter supports the all above options as @ref{commands}.
8779 @subsection Examples
8785 convolution="0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0"
8791 convolution="1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9"
8797 convolution="0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128"
8803 convolution="0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128"
8807 Apply laplacian edge detector which includes diagonals:
8809 convolution="1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0"
8815 convolution="-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2"
8821 Apply 2D convolution of video stream in frequency domain using second stream
8824 The filter accepts the following options:
8828 Set which planes to process.
8831 Set which impulse video frames will be processed, can be @var{first}
8832 or @var{all}. Default is @var{all}.
8835 The @code{convolve} filter also supports the @ref{framesync} options.
8839 Copy the input video source unchanged to the output. This is mainly useful for
8844 Video filtering on GPU using Apple's CoreImage API on OSX.
8846 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8847 processed by video hardware. However, software-based OpenGL implementations
8848 exist which means there is no guarantee for hardware processing. It depends on
8851 There are many filters and image generators provided by Apple that come with a
8852 large variety of options. The filter has to be referenced by its name along
8855 The coreimage filter accepts the following options:
8858 List all available filters and generators along with all their respective
8859 options as well as possible minimum and maximum values along with the default
8866 Specify all filters by their respective name and options.
8867 Use @var{list_filters} to determine all valid filter names and options.
8868 Numerical options are specified by a float value and are automatically clamped
8869 to their respective value range. Vector and color options have to be specified
8870 by a list of space separated float values. Character escaping has to be done.
8871 A special option name @code{default} is available to use default options for a
8874 It is required to specify either @code{default} or at least one of the filter options.
8875 All omitted options are used with their default values.
8876 The syntax of the filter string is as follows:
8878 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8882 Specify a rectangle where the output of the filter chain is copied into the
8883 input image. It is given by a list of space separated float values:
8885 output_rect=x\ y\ width\ height
8887 If not given, the output rectangle equals the dimensions of the input image.
8888 The output rectangle is automatically cropped at the borders of the input
8889 image. Negative values are valid for each component.
8891 output_rect=25\ 25\ 100\ 100
8895 Several filters can be chained for successive processing without GPU-HOST
8896 transfers allowing for fast processing of complex filter chains.
8897 Currently, only filters with zero (generators) or exactly one (filters) input
8898 image and one output image are supported. Also, transition filters are not yet
8901 Some filters generate output images with additional padding depending on the
8902 respective filter kernel. The padding is automatically removed to ensure the
8903 filter output has the same size as the input image.
8905 For image generators, the size of the output image is determined by the
8906 previous output image of the filter chain or the input image of the whole
8907 filterchain, respectively. The generators do not use the pixel information of
8908 this image to generate their output. However, the generated output is
8909 blended onto this image, resulting in partial or complete coverage of the
8912 The @ref{coreimagesrc} video source can be used for generating input images
8913 which are directly fed into the filter chain. By using it, providing input
8914 images by another video source or an input video is not required.
8916 @subsection Examples
8921 List all filters available:
8923 coreimage=list_filters=true
8927 Use the CIBoxBlur filter with default options to blur an image:
8929 coreimage=filter=CIBoxBlur@@default
8933 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8934 its center at 100x100 and a radius of 50 pixels:
8936 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8940 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8941 given as complete and escaped command-line for Apple's standard bash shell:
8943 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8949 Cover a rectangular object
8951 It accepts the following options:
8955 Filepath of the optional cover image, needs to be in yuv420.
8960 It accepts the following values:
8963 cover it by the supplied image
8965 cover it by interpolating the surrounding pixels
8968 Default value is @var{blur}.
8971 @subsection Examples
8975 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8977 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8983 Crop the input video to given dimensions.
8985 It accepts the following parameters:
8989 The width of the output video. It defaults to @code{iw}.
8990 This expression is evaluated only once during the filter
8991 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8994 The height of the output video. It defaults to @code{ih}.
8995 This expression is evaluated only once during the filter
8996 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8999 The horizontal position, in the input video, of the left edge of the output
9000 video. It defaults to @code{(in_w-out_w)/2}.
9001 This expression is evaluated per-frame.
9004 The vertical position, in the input video, of the top edge of the output video.
9005 It defaults to @code{(in_h-out_h)/2}.
9006 This expression is evaluated per-frame.
9009 If set to 1 will force the output display aspect ratio
9010 to be the same of the input, by changing the output sample aspect
9011 ratio. It defaults to 0.
9014 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
9015 width/height/x/y as specified and will not be rounded to nearest smaller value.
9019 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
9020 expressions containing the following constants:
9025 The computed values for @var{x} and @var{y}. They are evaluated for
9030 The input width and height.
9034 These are the same as @var{in_w} and @var{in_h}.
9038 The output (cropped) width and height.
9042 These are the same as @var{out_w} and @var{out_h}.
9045 same as @var{iw} / @var{ih}
9048 input sample aspect ratio
9051 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
9055 horizontal and vertical chroma subsample values. For example for the
9056 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9059 The number of the input frame, starting from 0.
9062 the position in the file of the input frame, NAN if unknown
9065 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
9069 The expression for @var{out_w} may depend on the value of @var{out_h},
9070 and the expression for @var{out_h} may depend on @var{out_w}, but they
9071 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
9072 evaluated after @var{out_w} and @var{out_h}.
9074 The @var{x} and @var{y} parameters specify the expressions for the
9075 position of the top-left corner of the output (non-cropped) area. They
9076 are evaluated for each frame. If the evaluated value is not valid, it
9077 is approximated to the nearest valid value.
9079 The expression for @var{x} may depend on @var{y}, and the expression
9080 for @var{y} may depend on @var{x}.
9082 @subsection Examples
9086 Crop area with size 100x100 at position (12,34).
9091 Using named options, the example above becomes:
9093 crop=w=100:h=100:x=12:y=34
9097 Crop the central input area with size 100x100:
9103 Crop the central input area with size 2/3 of the input video:
9105 crop=2/3*in_w:2/3*in_h
9109 Crop the input video central square:
9116 Delimit the rectangle with the top-left corner placed at position
9117 100:100 and the right-bottom corner corresponding to the right-bottom
9118 corner of the input image.
9120 crop=in_w-100:in_h-100:100:100
9124 Crop 10 pixels from the left and right borders, and 20 pixels from
9125 the top and bottom borders
9127 crop=in_w-2*10:in_h-2*20
9131 Keep only the bottom right quarter of the input image:
9133 crop=in_w/2:in_h/2:in_w/2:in_h/2
9137 Crop height for getting Greek harmony:
9139 crop=in_w:1/PHI*in_w
9143 Apply trembling effect:
9145 crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(n/10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(n/7)
9149 Apply erratic camera effect depending on timestamp:
9151 crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(t*10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(t*13)"
9155 Set x depending on the value of y:
9157 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
9161 @subsection Commands
9163 This filter supports the following commands:
9169 Set width/height of the output video and the horizontal/vertical position
9171 The command accepts the same syntax of the corresponding option.
9173 If the specified expression is not valid, it is kept at its current
9179 Auto-detect the crop size.
9181 It calculates the necessary cropping parameters and prints the
9182 recommended parameters via the logging system. The detected dimensions
9183 correspond to the non-black area of the input video.
9185 It accepts the following parameters:
9190 Set higher black value threshold, which can be optionally specified
9191 from nothing (0) to everything (255 for 8-bit based formats). An intensity
9192 value greater to the set value is considered non-black. It defaults to 24.
9193 You can also specify a value between 0.0 and 1.0 which will be scaled depending
9194 on the bitdepth of the pixel format.
9197 The value which the width/height should be divisible by. It defaults to
9198 16. The offset is automatically adjusted to center the video. Use 2 to
9199 get only even dimensions (needed for 4:2:2 video). 16 is best when
9200 encoding to most video codecs.
9203 Set the number of initial frames for which evaluation is skipped.
9204 Default is 2. Range is 0 to INT_MAX.
9206 @item reset_count, reset
9207 Set the counter that determines after how many frames cropdetect will
9208 reset the previously detected largest video area and start over to
9209 detect the current optimal crop area. Default value is 0.
9211 This can be useful when channel logos distort the video area. 0
9212 indicates 'never reset', and returns the largest area encountered during
9219 Delay video filtering until a given wallclock timestamp. The filter first
9220 passes on @option{preroll} amount of frames, then it buffers at most
9221 @option{buffer} amount of frames and waits for the cue. After reaching the cue
9222 it forwards the buffered frames and also any subsequent frames coming in its
9225 The filter can be used synchronize the output of multiple ffmpeg processes for
9226 realtime output devices like decklink. By putting the delay in the filtering
9227 chain and pre-buffering frames the process can pass on data to output almost
9228 immediately after the target wallclock timestamp is reached.
9230 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
9236 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
9239 The duration of content to pass on as preroll expressed in seconds. Default is 0.
9242 The maximum duration of content to buffer before waiting for the cue expressed
9243 in seconds. Default is 0.
9250 Apply color adjustments using curves.
9252 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
9253 component (red, green and blue) has its values defined by @var{N} key points
9254 tied from each other using a smooth curve. The x-axis represents the pixel
9255 values from the input frame, and the y-axis the new pixel values to be set for
9258 By default, a component curve is defined by the two points @var{(0;0)} and
9259 @var{(1;1)}. This creates a straight line where each original pixel value is
9260 "adjusted" to its own value, which means no change to the image.
9262 The filter allows you to redefine these two points and add some more. A new
9263 curve (using a natural cubic spline interpolation) will be define to pass
9264 smoothly through all these new coordinates. The new defined points needs to be
9265 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
9266 be in the @var{[0;1]} interval. If the computed curves happened to go outside
9267 the vector spaces, the values will be clipped accordingly.
9269 The filter accepts the following options:
9273 Select one of the available color presets. This option can be used in addition
9274 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
9275 options takes priority on the preset values.
9276 Available presets are:
9279 @item color_negative
9282 @item increase_contrast
9284 @item linear_contrast
9285 @item medium_contrast
9287 @item strong_contrast
9290 Default is @code{none}.
9292 Set the master key points. These points will define a second pass mapping. It
9293 is sometimes called a "luminance" or "value" mapping. It can be used with
9294 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
9295 post-processing LUT.
9297 Set the key points for the red component.
9299 Set the key points for the green component.
9301 Set the key points for the blue component.
9303 Set the key points for all components (not including master).
9304 Can be used in addition to the other key points component
9305 options. In this case, the unset component(s) will fallback on this
9306 @option{all} setting.
9308 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
9310 Save Gnuplot script of the curves in specified file.
9313 To avoid some filtergraph syntax conflicts, each key points list need to be
9314 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
9316 @subsection Commands
9318 This filter supports same @ref{commands} as options.
9320 @subsection Examples
9324 Increase slightly the middle level of blue:
9326 curves=blue='0/0 0.5/0.58 1/1'
9332 curves=r='0/0.11 .42/.51 1/0.95':g='0/0 0.50/0.48 1/1':b='0/0.22 .49/.44 1/0.8'
9334 Here we obtain the following coordinates for each components:
9337 @code{(0;0.11) (0.42;0.51) (1;0.95)}
9339 @code{(0;0) (0.50;0.48) (1;1)}
9341 @code{(0;0.22) (0.49;0.44) (1;0.80)}
9345 The previous example can also be achieved with the associated built-in preset:
9347 curves=preset=vintage
9357 Use a Photoshop preset and redefine the points of the green component:
9359 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
9363 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
9364 and @command{gnuplot}:
9366 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
9367 gnuplot -p /tmp/curves.plt
9373 Video data analysis filter.
9375 This filter shows hexadecimal pixel values of part of video.
9377 The filter accepts the following options:
9381 Set output video size.
9384 Set x offset from where to pick pixels.
9387 Set y offset from where to pick pixels.
9390 Set scope mode, can be one of the following:
9393 Draw hexadecimal pixel values with white color on black background.
9396 Draw hexadecimal pixel values with input video pixel color on black
9400 Draw hexadecimal pixel values on color background picked from input video,
9401 the text color is picked in such way so its always visible.
9405 Draw rows and columns numbers on left and top of video.
9408 Set background opacity.
9411 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
9414 Set pixel components to display. By default all pixel components are displayed.
9417 @subsection Commands
9419 This filter supports same @ref{commands} as options excluding @code{size} option.
9422 Apply Directional blur filter.
9424 The filter accepts the following options:
9428 Set angle of directional blur. Default is @code{45}.
9431 Set radius of directional blur. Default is @code{5}.
9434 Set which planes to filter. By default all planes are filtered.
9437 @subsection Commands
9438 This filter supports same @ref{commands} as options.
9439 The command accepts the same syntax of the corresponding option.
9441 If the specified expression is not valid, it is kept at its current
9446 Denoise frames using 2D DCT (frequency domain filtering).
9448 This filter is not designed for real time.
9450 The filter accepts the following options:
9454 Set the noise sigma constant.
9456 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
9457 coefficient (absolute value) below this threshold with be dropped.
9459 If you need a more advanced filtering, see @option{expr}.
9461 Default is @code{0}.
9464 Set number overlapping pixels for each block. Since the filter can be slow, you
9465 may want to reduce this value, at the cost of a less effective filter and the
9466 risk of various artefacts.
9468 If the overlapping value doesn't permit processing the whole input width or
9469 height, a warning will be displayed and according borders won't be denoised.
9471 Default value is @var{blocksize}-1, which is the best possible setting.
9474 Set the coefficient factor expression.
9476 For each coefficient of a DCT block, this expression will be evaluated as a
9477 multiplier value for the coefficient.
9479 If this is option is set, the @option{sigma} option will be ignored.
9481 The absolute value of the coefficient can be accessed through the @var{c}
9485 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
9486 @var{blocksize}, which is the width and height of the processed blocks.
9488 The default value is @var{3} (8x8) and can be raised to @var{4} for a
9489 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
9490 on the speed processing. Also, a larger block size does not necessarily means a
9494 @subsection Examples
9496 Apply a denoise with a @option{sigma} of @code{4.5}:
9501 The same operation can be achieved using the expression system:
9503 dctdnoiz=e='gte(c, 4.5*3)'
9506 Violent denoise using a block size of @code{16x16}:
9513 Remove banding artifacts from input video.
9514 It works by replacing banded pixels with average value of referenced pixels.
9516 The filter accepts the following options:
9523 Set banding detection threshold for each plane. Default is 0.02.
9524 Valid range is 0.00003 to 0.5.
9525 If difference between current pixel and reference pixel is less than threshold,
9526 it will be considered as banded.
9529 Banding detection range in pixels. Default is 16. If positive, random number
9530 in range 0 to set value will be used. If negative, exact absolute value
9532 The range defines square of four pixels around current pixel.
9535 Set direction in radians from which four pixel will be compared. If positive,
9536 random direction from 0 to set direction will be picked. If negative, exact of
9537 absolute value will be picked. For example direction 0, -PI or -2*PI radians
9538 will pick only pixels on same row and -PI/2 will pick only pixels on same
9542 If enabled, current pixel is compared with average value of all four
9543 surrounding pixels. The default is enabled. If disabled current pixel is
9544 compared with all four surrounding pixels. The pixel is considered banded
9545 if only all four differences with surrounding pixels are less than threshold.
9548 If enabled, current pixel is changed if and only if all pixel components are banded,
9549 e.g. banding detection threshold is triggered for all color components.
9550 The default is disabled.
9553 @subsection Commands
9555 This filter supports the all above options as @ref{commands}.
9559 Remove blocking artifacts from input video.
9561 The filter accepts the following options:
9565 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
9566 This controls what kind of deblocking is applied.
9569 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
9575 Set blocking detection thresholds. Allowed range is 0 to 1.
9576 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
9577 Using higher threshold gives more deblocking strength.
9578 Setting @var{alpha} controls threshold detection at exact edge of block.
9579 Remaining options controls threshold detection near the edge. Each one for
9580 below/above or left/right. Setting any of those to @var{0} disables
9584 Set planes to filter. Default is to filter all available planes.
9587 @subsection Examples
9591 Deblock using weak filter and block size of 4 pixels.
9593 deblock=filter=weak:block=4
9597 Deblock using strong filter, block size of 4 pixels and custom thresholds for
9598 deblocking more edges.
9600 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
9604 Similar as above, but filter only first plane.
9606 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
9610 Similar as above, but filter only second and third plane.
9612 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
9616 @subsection Commands
9618 This filter supports the all above options as @ref{commands}.
9623 Drop duplicated frames at regular intervals.
9625 The filter accepts the following options:
9629 Set the number of frames from which one will be dropped. Setting this to
9630 @var{N} means one frame in every batch of @var{N} frames will be dropped.
9631 Default is @code{5}.
9634 Set the threshold for duplicate detection. If the difference metric for a frame
9635 is less than or equal to this value, then it is declared as duplicate. Default
9639 Set scene change threshold. Default is @code{15}.
9643 Set the size of the x and y-axis blocks used during metric calculations.
9644 Larger blocks give better noise suppression, but also give worse detection of
9645 small movements. Must be a power of two. Default is @code{32}.
9648 Mark main input as a pre-processed input and activate clean source input
9649 stream. This allows the input to be pre-processed with various filters to help
9650 the metrics calculation while keeping the frame selection lossless. When set to
9651 @code{1}, the first stream is for the pre-processed input, and the second
9652 stream is the clean source from where the kept frames are chosen. Default is
9656 Set whether or not chroma is considered in the metric calculations. Default is
9662 Apply 2D deconvolution of video stream in frequency domain using second stream
9665 The filter accepts the following options:
9669 Set which planes to process.
9672 Set which impulse video frames will be processed, can be @var{first}
9673 or @var{all}. Default is @var{all}.
9676 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
9677 and height are not same and not power of 2 or if stream prior to convolving
9681 The @code{deconvolve} filter also supports the @ref{framesync} options.
9685 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
9687 It accepts the following options:
9691 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
9692 @var{rainbows} for cross-color reduction.
9695 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
9698 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
9701 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
9704 Set temporal chroma threshold. Lower values increases reduction of cross-color.
9709 Apply deflate effect to the video.
9711 This filter replaces the pixel by the local(3x3) average by taking into account
9712 only values lower than the pixel.
9714 It accepts the following options:
9721 Limit the maximum change for each plane, default is 65535.
9722 If 0, plane will remain unchanged.
9725 @subsection Commands
9727 This filter supports the all above options as @ref{commands}.
9731 Remove temporal frame luminance variations.
9733 It accepts the following options:
9737 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
9740 Set averaging mode to smooth temporal luminance variations.
9742 Available values are:
9767 Do not actually modify frame. Useful when one only wants metadata.
9772 Remove judder produced by partially interlaced telecined content.
9774 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
9775 source was partially telecined content then the output of @code{pullup,dejudder}
9776 will have a variable frame rate. May change the recorded frame rate of the
9777 container. Aside from that change, this filter will not affect constant frame
9780 The option available in this filter is:
9784 Specify the length of the window over which the judder repeats.
9786 Accepts any integer greater than 1. Useful values are:
9790 If the original was telecined from 24 to 30 fps (Film to NTSC).
9793 If the original was telecined from 25 to 30 fps (PAL to NTSC).
9796 If a mixture of the two.
9799 The default is @samp{4}.
9804 Suppress a TV station logo by a simple interpolation of the surrounding
9805 pixels. Just set a rectangle covering the logo and watch it disappear
9806 (and sometimes something even uglier appear - your mileage may vary).
9808 It accepts the following parameters:
9813 Specify the top left corner coordinates of the logo. They must be
9818 Specify the width and height of the logo to clear. They must be
9822 Specify the thickness of the fuzzy edge of the rectangle (added to
9823 @var{w} and @var{h}). The default value is 1. This option is
9824 deprecated, setting higher values should no longer be necessary and
9828 When set to 1, a green rectangle is drawn on the screen to simplify
9829 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9830 The default value is 0.
9832 The rectangle is drawn on the outermost pixels which will be (partly)
9833 replaced with interpolated values. The values of the next pixels
9834 immediately outside this rectangle in each direction will be used to
9835 compute the interpolated pixel values inside the rectangle.
9839 @subsection Examples
9843 Set a rectangle covering the area with top left corner coordinates 0,0
9844 and size 100x77, and a band of size 10:
9846 delogo=x=0:y=0:w=100:h=77:band=10
9854 Remove the rain in the input image/video by applying the derain methods based on
9855 convolutional neural networks. Supported models:
9859 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9860 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9863 Training as well as model generation scripts are provided in
9864 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9866 Native model files (.model) can be generated from TensorFlow model
9867 files (.pb) by using tools/python/convert.py
9869 The filter accepts the following options:
9873 Specify which filter to use. This option accepts the following values:
9877 Derain filter. To conduct derain filter, you need to use a derain model.
9880 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9882 Default value is @samp{derain}.
9885 Specify which DNN backend to use for model loading and execution. This option accepts
9886 the following values:
9890 Native implementation of DNN loading and execution.
9893 TensorFlow backend. To enable this backend you
9894 need to install the TensorFlow for C library (see
9895 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9896 @code{--enable-libtensorflow}
9898 Default value is @samp{native}.
9901 Set path to model file specifying network architecture and its parameters.
9902 Note that different backends use different file formats. TensorFlow and native
9903 backend can load files for only its format.
9906 It can also be finished with @ref{dnn_processing} filter.
9910 Attempt to fix small changes in horizontal and/or vertical shift. This
9911 filter helps remove camera shake from hand-holding a camera, bumping a
9912 tripod, moving on a vehicle, etc.
9914 The filter accepts the following options:
9922 Specify a rectangular area where to limit the search for motion
9924 If desired the search for motion vectors can be limited to a
9925 rectangular area of the frame defined by its top left corner, width
9926 and height. These parameters have the same meaning as the drawbox
9927 filter which can be used to visualise the position of the bounding
9930 This is useful when simultaneous movement of subjects within the frame
9931 might be confused for camera motion by the motion vector search.
9933 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9934 then the full frame is used. This allows later options to be set
9935 without specifying the bounding box for the motion vector search.
9937 Default - search the whole frame.
9941 Specify the maximum extent of movement in x and y directions in the
9942 range 0-64 pixels. Default 16.
9945 Specify how to generate pixels to fill blanks at the edge of the
9946 frame. Available values are:
9949 Fill zeroes at blank locations
9951 Original image at blank locations
9953 Extruded edge value at blank locations
9955 Mirrored edge at blank locations
9957 Default value is @samp{mirror}.
9960 Specify the blocksize to use for motion search. Range 4-128 pixels,
9964 Specify the contrast threshold for blocks. Only blocks with more than
9965 the specified contrast (difference between darkest and lightest
9966 pixels) will be considered. Range 1-255, default 125.
9969 Specify the search strategy. Available values are:
9972 Set exhaustive search
9974 Set less exhaustive search.
9976 Default value is @samp{exhaustive}.
9979 If set then a detailed log of the motion search is written to the
9986 Remove unwanted contamination of foreground colors, caused by reflected color of
9987 greenscreen or bluescreen.
9989 This filter accepts the following options:
9993 Set what type of despill to use.
9996 Set how spillmap will be generated.
9999 Set how much to get rid of still remaining spill.
10002 Controls amount of red in spill area.
10005 Controls amount of green in spill area.
10006 Should be -1 for greenscreen.
10009 Controls amount of blue in spill area.
10010 Should be -1 for bluescreen.
10013 Controls brightness of spill area, preserving colors.
10016 Modify alpha from generated spillmap.
10019 @subsection Commands
10021 This filter supports the all above options as @ref{commands}.
10023 @section detelecine
10025 Apply an exact inverse of the telecine operation. It requires a predefined
10026 pattern specified using the pattern option which must be the same as that passed
10027 to the telecine filter.
10029 This filter accepts the following options:
10038 The default value is @code{top}.
10042 A string of numbers representing the pulldown pattern you wish to apply.
10043 The default value is @code{23}.
10046 A number representing position of the first frame with respect to the telecine
10047 pattern. This is to be used if the stream is cut. The default value is @code{0}.
10052 Apply dilation effect to the video.
10054 This filter replaces the pixel by the local(3x3) maximum.
10056 It accepts the following options:
10063 Limit the maximum change for each plane, default is 65535.
10064 If 0, plane will remain unchanged.
10067 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
10070 Flags to local 3x3 coordinates maps like this:
10077 @subsection Commands
10079 This filter supports the all above options as @ref{commands}.
10083 Displace pixels as indicated by second and third input stream.
10085 It takes three input streams and outputs one stream, the first input is the
10086 source, and second and third input are displacement maps.
10088 The second input specifies how much to displace pixels along the
10089 x-axis, while the third input specifies how much to displace pixels
10091 If one of displacement map streams terminates, last frame from that
10092 displacement map will be used.
10094 Note that once generated, displacements maps can be reused over and over again.
10096 A description of the accepted options follows.
10100 Set displace behavior for pixels that are out of range.
10102 Available values are:
10105 Missing pixels are replaced by black pixels.
10108 Adjacent pixels will spread out to replace missing pixels.
10111 Out of range pixels are wrapped so they point to pixels of other side.
10114 Out of range pixels will be replaced with mirrored pixels.
10116 Default is @samp{smear}.
10120 @subsection Examples
10124 Add ripple effect to rgb input of video size hd720:
10126 ffmpeg -i INPUT -f lavfi -i nullsrc=s=hd720,lutrgb=128:128:128 -f lavfi -i nullsrc=s=hd720,geq='r=128+30*sin(2*PI*X/400+T):g=128+30*sin(2*PI*X/400+T):b=128+30*sin(2*PI*X/400+T)' -lavfi '[0][1][2]displace' OUTPUT
10130 Add wave effect to rgb input of video size hd720:
10132 ffmpeg -i INPUT -f lavfi -i nullsrc=hd720,geq='r=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):g=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):b=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T))' -lavfi '[1]split[x][y],[0][x][y]displace' OUTPUT
10136 @anchor{dnn_processing}
10137 @section dnn_processing
10139 Do image processing with deep neural networks. It works together with another filter
10140 which converts the pixel format of the Frame to what the dnn network requires.
10142 The filter accepts the following options:
10146 Specify which DNN backend to use for model loading and execution. This option accepts
10147 the following values:
10151 Native implementation of DNN loading and execution.
10154 TensorFlow backend. To enable this backend you
10155 need to install the TensorFlow for C library (see
10156 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
10157 @code{--enable-libtensorflow}
10160 OpenVINO backend. To enable this backend you
10161 need to build and install the OpenVINO for C library (see
10162 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
10163 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
10164 be needed if the header files and libraries are not installed into system path)
10168 Default value is @samp{native}.
10171 Set path to model file specifying network architecture and its parameters.
10172 Note that different backends use different file formats. TensorFlow, OpenVINO and native
10173 backend can load files for only its format.
10175 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
10178 Set the input name of the dnn network.
10181 Set the output name of the dnn network.
10184 use DNN async execution if set (default: set),
10185 roll back to sync execution if the backend does not support async.
10189 @subsection Examples
10193 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
10195 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
10199 Halve the pixel value of the frame with format gray32f:
10201 ffmpeg -i input.jpg -vf format=grayf32,dnn_processing=model=halve_gray_float.model:input=dnn_in:output=dnn_out:dnn_backend=native -y out.native.png
10205 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
10207 ./ffmpeg -i 480p.jpg -vf format=yuv420p,scale=w=iw*2:h=ih*2,dnn_processing=dnn_backend=tensorflow:model=srcnn.pb:input=x:output=y -y srcnn.jpg
10211 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
10213 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
10220 Draw a colored box on the input image.
10222 It accepts the following parameters:
10227 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
10231 The expressions which specify the width and height of the box; if 0 they are interpreted as
10232 the input width and height. It defaults to 0.
10235 Specify the color of the box to write. For the general syntax of this option,
10236 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10237 value @code{invert} is used, the box edge color is the same as the
10238 video with inverted luma.
10241 The expression which sets the thickness of the box edge.
10242 A value of @code{fill} will create a filled box. Default value is @code{3}.
10244 See below for the list of accepted constants.
10247 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
10248 will overwrite the video's color and alpha pixels.
10249 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
10252 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10253 following constants:
10257 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10261 horizontal and vertical chroma subsample values. For example for the
10262 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10266 The input width and height.
10269 The input sample aspect ratio.
10273 The x and y offset coordinates where the box is drawn.
10277 The width and height of the drawn box.
10280 The thickness of the drawn box.
10282 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10283 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10287 @subsection Examples
10291 Draw a black box around the edge of the input image:
10297 Draw a box with color red and an opacity of 50%:
10299 drawbox=10:20:200:60:red@@0.5
10302 The previous example can be specified as:
10304 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
10308 Fill the box with pink color:
10310 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
10314 Draw a 2-pixel red 2.40:1 mask:
10316 drawbox=x=-t:y=0.5*(ih-iw/2.4)-t:w=iw+t*2:h=iw/2.4+t*2:t=2:c=red
10320 @subsection Commands
10321 This filter supports same commands as options.
10322 The command accepts the same syntax of the corresponding option.
10324 If the specified expression is not valid, it is kept at its current
10329 Draw a graph using input video metadata.
10331 It accepts the following parameters:
10335 Set 1st frame metadata key from which metadata values will be used to draw a graph.
10338 Set 1st foreground color expression.
10341 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
10344 Set 2nd foreground color expression.
10347 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
10350 Set 3rd foreground color expression.
10353 Set 4th frame metadata key from which metadata values will be used to draw a graph.
10356 Set 4th foreground color expression.
10359 Set minimal value of metadata value.
10362 Set maximal value of metadata value.
10365 Set graph background color. Default is white.
10370 Available values for mode is:
10377 Default is @code{line}.
10382 Available values for slide is:
10385 Draw new frame when right border is reached.
10388 Replace old columns with new ones.
10391 Scroll from right to left.
10394 Scroll from left to right.
10397 Draw single picture.
10400 Default is @code{frame}.
10403 Set size of graph video. For the syntax of this option, check the
10404 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
10405 The default value is @code{900x256}.
10408 Set the output frame rate. Default value is @code{25}.
10410 The foreground color expressions can use the following variables:
10413 Minimal value of metadata value.
10416 Maximal value of metadata value.
10419 Current metadata key value.
10422 The color is defined as 0xAABBGGRR.
10425 Example using metadata from @ref{signalstats} filter:
10427 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
10430 Example using metadata from @ref{ebur128} filter:
10432 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
10437 Draw a grid on the input image.
10439 It accepts the following parameters:
10444 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
10448 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
10449 input width and height, respectively, minus @code{thickness}, so image gets
10450 framed. Default to 0.
10453 Specify the color of the grid. For the general syntax of this option,
10454 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10455 value @code{invert} is used, the grid color is the same as the
10456 video with inverted luma.
10459 The expression which sets the thickness of the grid line. Default value is @code{1}.
10461 See below for the list of accepted constants.
10464 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
10465 will overwrite the video's color and alpha pixels.
10466 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
10469 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10470 following constants:
10474 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10478 horizontal and vertical chroma subsample values. For example for the
10479 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10483 The input grid cell width and height.
10486 The input sample aspect ratio.
10490 The x and y coordinates of some point of grid intersection (meant to configure offset).
10494 The width and height of the drawn cell.
10497 The thickness of the drawn cell.
10499 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10500 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10504 @subsection Examples
10508 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
10510 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
10514 Draw a white 3x3 grid with an opacity of 50%:
10516 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
10520 @subsection Commands
10521 This filter supports same commands as options.
10522 The command accepts the same syntax of the corresponding option.
10524 If the specified expression is not valid, it is kept at its current
10530 Draw a text string or text from a specified file on top of a video, using the
10531 libfreetype library.
10533 To enable compilation of this filter, you need to configure FFmpeg with
10534 @code{--enable-libfreetype}.
10535 To enable default font fallback and the @var{font} option you need to
10536 configure FFmpeg with @code{--enable-libfontconfig}.
10537 To enable the @var{text_shaping} option, you need to configure FFmpeg with
10538 @code{--enable-libfribidi}.
10542 It accepts the following parameters:
10547 Used to draw a box around text using the background color.
10548 The value must be either 1 (enable) or 0 (disable).
10549 The default value of @var{box} is 0.
10552 Set the width of the border to be drawn around the box using @var{boxcolor}.
10553 The default value of @var{boxborderw} is 0.
10556 The color to be used for drawing box around text. For the syntax of this
10557 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10559 The default value of @var{boxcolor} is "white".
10562 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
10563 The default value of @var{line_spacing} is 0.
10566 Set the width of the border to be drawn around the text using @var{bordercolor}.
10567 The default value of @var{borderw} is 0.
10570 Set the color to be used for drawing border around text. For the syntax of this
10571 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10573 The default value of @var{bordercolor} is "black".
10576 Select how the @var{text} is expanded. Can be either @code{none},
10577 @code{strftime} (deprecated) or
10578 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
10582 Set a start time for the count. Value is in microseconds. Only applied
10583 in the deprecated strftime expansion mode. To emulate in normal expansion
10584 mode use the @code{pts} function, supplying the start time (in seconds)
10585 as the second argument.
10588 If true, check and fix text coords to avoid clipping.
10591 The color to be used for drawing fonts. For the syntax of this option, check
10592 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10594 The default value of @var{fontcolor} is "black".
10596 @item fontcolor_expr
10597 String which is expanded the same way as @var{text} to obtain dynamic
10598 @var{fontcolor} value. By default this option has empty value and is not
10599 processed. When this option is set, it overrides @var{fontcolor} option.
10602 The font family to be used for drawing text. By default Sans.
10605 The font file to be used for drawing text. The path must be included.
10606 This parameter is mandatory if the fontconfig support is disabled.
10609 Draw the text applying alpha blending. The value can
10610 be a number between 0.0 and 1.0.
10611 The expression accepts the same variables @var{x, y} as well.
10612 The default value is 1.
10613 Please see @var{fontcolor_expr}.
10616 The font size to be used for drawing text.
10617 The default value of @var{fontsize} is 16.
10620 If set to 1, attempt to shape the text (for example, reverse the order of
10621 right-to-left text and join Arabic characters) before drawing it.
10622 Otherwise, just draw the text exactly as given.
10623 By default 1 (if supported).
10625 @item ft_load_flags
10626 The flags to be used for loading the fonts.
10628 The flags map the corresponding flags supported by libfreetype, and are
10629 a combination of the following values:
10636 @item vertical_layout
10637 @item force_autohint
10640 @item ignore_global_advance_width
10642 @item ignore_transform
10644 @item linear_design
10648 Default value is "default".
10650 For more information consult the documentation for the FT_LOAD_*
10654 The color to be used for drawing a shadow behind the drawn text. For the
10655 syntax of this option, check the @ref{color syntax,,"Color" section in the
10656 ffmpeg-utils manual,ffmpeg-utils}.
10658 The default value of @var{shadowcolor} is "black".
10662 The x and y offsets for the text shadow position with respect to the
10663 position of the text. They can be either positive or negative
10664 values. The default value for both is "0".
10667 The starting frame number for the n/frame_num variable. The default value
10671 The size in number of spaces to use for rendering the tab.
10672 Default value is 4.
10675 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
10676 format. It can be used with or without text parameter. @var{timecode_rate}
10677 option must be specified.
10679 @item timecode_rate, rate, r
10680 Set the timecode frame rate (timecode only). Value will be rounded to nearest
10681 integer. Minimum value is "1".
10682 Drop-frame timecode is supported for frame rates 30 & 60.
10685 If set to 1, the output of the timecode option will wrap around at 24 hours.
10686 Default is 0 (disabled).
10689 The text string to be drawn. The text must be a sequence of UTF-8
10690 encoded characters.
10691 This parameter is mandatory if no file is specified with the parameter
10695 A text file containing text to be drawn. The text must be a sequence
10696 of UTF-8 encoded characters.
10698 This parameter is mandatory if no text string is specified with the
10699 parameter @var{text}.
10701 If both @var{text} and @var{textfile} are specified, an error is thrown.
10704 If set to 1, the @var{textfile} will be reloaded before each frame.
10705 Be sure to update it atomically, or it may be read partially, or even fail.
10709 The expressions which specify the offsets where text will be drawn
10710 within the video frame. They are relative to the top/left border of the
10713 The default value of @var{x} and @var{y} is "0".
10715 See below for the list of accepted constants and functions.
10718 The parameters for @var{x} and @var{y} are expressions containing the
10719 following constants and functions:
10723 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
10727 horizontal and vertical chroma subsample values. For example for the
10728 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10731 the height of each text line
10739 @item max_glyph_a, ascent
10740 the maximum distance from the baseline to the highest/upper grid
10741 coordinate used to place a glyph outline point, for all the rendered
10743 It is a positive value, due to the grid's orientation with the Y axis
10746 @item max_glyph_d, descent
10747 the maximum distance from the baseline to the lowest grid coordinate
10748 used to place a glyph outline point, for all the rendered glyphs.
10749 This is a negative value, due to the grid's orientation, with the Y axis
10753 maximum glyph height, that is the maximum height for all the glyphs
10754 contained in the rendered text, it is equivalent to @var{ascent} -
10758 maximum glyph width, that is the maximum width for all the glyphs
10759 contained in the rendered text
10762 the number of input frame, starting from 0
10764 @item rand(min, max)
10765 return a random number included between @var{min} and @var{max}
10768 The input sample aspect ratio.
10771 timestamp expressed in seconds, NAN if the input timestamp is unknown
10774 the height of the rendered text
10777 the width of the rendered text
10781 the x and y offset coordinates where the text is drawn.
10783 These parameters allow the @var{x} and @var{y} expressions to refer
10784 to each other, so you can for example specify @code{y=x/dar}.
10787 A one character description of the current frame's picture type.
10790 The current packet's position in the input file or stream
10791 (in bytes, from the start of the input). A value of -1 indicates
10792 this info is not available.
10795 The current packet's duration, in seconds.
10798 The current packet's size (in bytes).
10801 @anchor{drawtext_expansion}
10802 @subsection Text expansion
10804 If @option{expansion} is set to @code{strftime},
10805 the filter recognizes strftime() sequences in the provided text and
10806 expands them accordingly. Check the documentation of strftime(). This
10807 feature is deprecated.
10809 If @option{expansion} is set to @code{none}, the text is printed verbatim.
10811 If @option{expansion} is set to @code{normal} (which is the default),
10812 the following expansion mechanism is used.
10814 The backslash character @samp{\}, followed by any character, always expands to
10815 the second character.
10817 Sequences of the form @code{%@{...@}} are expanded. The text between the
10818 braces is a function name, possibly followed by arguments separated by ':'.
10819 If the arguments contain special characters or delimiters (':' or '@}'),
10820 they should be escaped.
10822 Note that they probably must also be escaped as the value for the
10823 @option{text} option in the filter argument string and as the filter
10824 argument in the filtergraph description, and possibly also for the shell,
10825 that makes up to four levels of escaping; using a text file avoids these
10828 The following functions are available:
10833 The expression evaluation result.
10835 It must take one argument specifying the expression to be evaluated,
10836 which accepts the same constants and functions as the @var{x} and
10837 @var{y} values. Note that not all constants should be used, for
10838 example the text size is not known when evaluating the expression, so
10839 the constants @var{text_w} and @var{text_h} will have an undefined
10842 @item expr_int_format, eif
10843 Evaluate the expression's value and output as formatted integer.
10845 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10846 The second argument specifies the output format. Allowed values are @samp{x},
10847 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10848 @code{printf} function.
10849 The third parameter is optional and sets the number of positions taken by the output.
10850 It can be used to add padding with zeros from the left.
10853 The time at which the filter is running, expressed in UTC.
10854 It can accept an argument: a strftime() format string.
10857 The time at which the filter is running, expressed in the local time zone.
10858 It can accept an argument: a strftime() format string.
10861 Frame metadata. Takes one or two arguments.
10863 The first argument is mandatory and specifies the metadata key.
10865 The second argument is optional and specifies a default value, used when the
10866 metadata key is not found or empty.
10868 Available metadata can be identified by inspecting entries
10869 starting with TAG included within each frame section
10870 printed by running @code{ffprobe -show_frames}.
10872 String metadata generated in filters leading to
10873 the drawtext filter are also available.
10876 The frame number, starting from 0.
10879 A one character description of the current picture type.
10882 The timestamp of the current frame.
10883 It can take up to three arguments.
10885 The first argument is the format of the timestamp; it defaults to @code{flt}
10886 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10887 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10888 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10889 @code{localtime} stands for the timestamp of the frame formatted as
10890 local time zone time.
10892 The second argument is an offset added to the timestamp.
10894 If the format is set to @code{hms}, a third argument @code{24HH} may be
10895 supplied to present the hour part of the formatted timestamp in 24h format
10898 If the format is set to @code{localtime} or @code{gmtime},
10899 a third argument may be supplied: a strftime() format string.
10900 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10903 @subsection Commands
10905 This filter supports altering parameters via commands:
10908 Alter existing filter parameters.
10910 Syntax for the argument is the same as for filter invocation, e.g.
10913 fontsize=56:fontcolor=green:text='Hello World'
10916 Full filter invocation with sendcmd would look like this:
10919 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10923 If the entire argument can't be parsed or applied as valid values then the filter will
10924 continue with its existing parameters.
10926 @subsection Examples
10930 Draw "Test Text" with font FreeSerif, using the default values for the
10931 optional parameters.
10934 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10938 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10939 and y=50 (counting from the top-left corner of the screen), text is
10940 yellow with a red box around it. Both the text and the box have an
10944 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10945 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10948 Note that the double quotes are not necessary if spaces are not used
10949 within the parameter list.
10952 Show the text at the center of the video frame:
10954 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10958 Show the text at a random position, switching to a new position every 30 seconds:
10960 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=if(eq(mod(t\,30)\,0)\,rand(0\,(w-text_w))\,x):y=if(eq(mod(t\,30)\,0)\,rand(0\,(h-text_h))\,y)"
10964 Show a text line sliding from right to left in the last row of the video
10965 frame. The file @file{LONG_LINE} is assumed to contain a single line
10968 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10972 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10974 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10978 Draw a single green letter "g", at the center of the input video.
10979 The glyph baseline is placed at half screen height.
10981 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10985 Show text for 1 second every 3 seconds:
10987 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10991 Use fontconfig to set the font. Note that the colons need to be escaped.
10993 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10997 Draw "Test Text" with font size dependent on height of the video.
10999 drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
11003 Print the date of a real-time encoding (see strftime(3)):
11005 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
11009 Show text fading in and out (appearing/disappearing):
11012 DS=1.0 # display start
11013 DE=10.0 # display end
11014 FID=1.5 # fade in duration
11015 FOD=5 # fade out duration
11016 ffplay -f lavfi "color,drawtext=text=TEST:fontsize=50:fontfile=FreeSerif.ttf:fontcolor_expr=ff0000%@{eif\\\\: clip(255*(1*between(t\\, $DS + $FID\\, $DE - $FOD) + ((t - $DS)/$FID)*between(t\\, $DS\\, $DS + $FID) + (-(t - $DE)/$FOD)*between(t\\, $DE - $FOD\\, $DE) )\\, 0\\, 255) \\\\: x\\\\: 2 @}"
11020 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
11021 and the @option{fontsize} value are included in the @option{y} offset.
11023 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
11024 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
11028 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
11029 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
11030 must have option @option{-export_path_metadata 1} for the special metadata fields
11031 to be available for filters.
11033 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
11038 For more information about libfreetype, check:
11039 @url{http://www.freetype.org/}.
11041 For more information about fontconfig, check:
11042 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
11044 For more information about libfribidi, check:
11045 @url{http://fribidi.org/}.
11047 @section edgedetect
11049 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
11051 The filter accepts the following options:
11056 Set low and high threshold values used by the Canny thresholding
11059 The high threshold selects the "strong" edge pixels, which are then
11060 connected through 8-connectivity with the "weak" edge pixels selected
11061 by the low threshold.
11063 @var{low} and @var{high} threshold values must be chosen in the range
11064 [0,1], and @var{low} should be lesser or equal to @var{high}.
11066 Default value for @var{low} is @code{20/255}, and default value for @var{high}
11070 Define the drawing mode.
11074 Draw white/gray wires on black background.
11077 Mix the colors to create a paint/cartoon effect.
11080 Apply Canny edge detector on all selected planes.
11082 Default value is @var{wires}.
11085 Select planes for filtering. By default all available planes are filtered.
11088 @subsection Examples
11092 Standard edge detection with custom values for the hysteresis thresholding:
11094 edgedetect=low=0.1:high=0.4
11098 Painting effect without thresholding:
11100 edgedetect=mode=colormix:high=0
11106 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
11108 For each input image, the filter will compute the optimal mapping from
11109 the input to the output given the codebook length, that is the number
11110 of distinct output colors.
11112 This filter accepts the following options.
11115 @item codebook_length, l
11116 Set codebook length. The value must be a positive integer, and
11117 represents the number of distinct output colors. Default value is 256.
11120 Set the maximum number of iterations to apply for computing the optimal
11121 mapping. The higher the value the better the result and the higher the
11122 computation time. Default value is 1.
11125 Set a random seed, must be an integer included between 0 and
11126 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
11127 will try to use a good random seed on a best effort basis.
11130 Set pal8 output pixel format. This option does not work with codebook
11131 length greater than 256.
11136 Measure graylevel entropy in histogram of color channels of video frames.
11138 It accepts the following parameters:
11142 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
11144 @var{diff} mode measures entropy of histogram delta values, absolute differences
11145 between neighbour histogram values.
11149 Apply the EPX magnification filter which is designed for pixel art.
11151 It accepts the following option:
11155 Set the scaling dimension: @code{2} for @code{2xEPX}, @code{3} for
11157 Default is @code{3}.
11161 Set brightness, contrast, saturation and approximate gamma adjustment.
11163 The filter accepts the following options:
11167 Set the contrast expression. The value must be a float value in range
11168 @code{-1000.0} to @code{1000.0}. The default value is "1".
11171 Set the brightness expression. The value must be a float value in
11172 range @code{-1.0} to @code{1.0}. The default value is "0".
11175 Set the saturation expression. The value must be a float in
11176 range @code{0.0} to @code{3.0}. The default value is "1".
11179 Set the gamma expression. The value must be a float in range
11180 @code{0.1} to @code{10.0}. The default value is "1".
11183 Set the gamma expression for red. The value must be a float in
11184 range @code{0.1} to @code{10.0}. The default value is "1".
11187 Set the gamma expression for green. The value must be a float in range
11188 @code{0.1} to @code{10.0}. The default value is "1".
11191 Set the gamma expression for blue. The value must be a float in range
11192 @code{0.1} to @code{10.0}. The default value is "1".
11195 Set the gamma weight expression. It can be used to reduce the effect
11196 of a high gamma value on bright image areas, e.g. keep them from
11197 getting overamplified and just plain white. The value must be a float
11198 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
11199 gamma correction all the way down while @code{1.0} leaves it at its
11200 full strength. Default is "1".
11203 Set when the expressions for brightness, contrast, saturation and
11204 gamma expressions are evaluated.
11206 It accepts the following values:
11209 only evaluate expressions once during the filter initialization or
11210 when a command is processed
11213 evaluate expressions for each incoming frame
11216 Default value is @samp{init}.
11219 The expressions accept the following parameters:
11222 frame count of the input frame starting from 0
11225 byte position of the corresponding packet in the input file, NAN if
11229 frame rate of the input video, NAN if the input frame rate is unknown
11232 timestamp expressed in seconds, NAN if the input timestamp is unknown
11235 @subsection Commands
11236 The filter supports the following commands:
11240 Set the contrast expression.
11243 Set the brightness expression.
11246 Set the saturation expression.
11249 Set the gamma expression.
11252 Set the gamma_r expression.
11255 Set gamma_g expression.
11258 Set gamma_b expression.
11261 Set gamma_weight expression.
11263 The command accepts the same syntax of the corresponding option.
11265 If the specified expression is not valid, it is kept at its current
11272 Apply erosion effect to the video.
11274 This filter replaces the pixel by the local(3x3) minimum.
11276 It accepts the following options:
11283 Limit the maximum change for each plane, default is 65535.
11284 If 0, plane will remain unchanged.
11287 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
11290 Flags to local 3x3 coordinates maps like this:
11297 @subsection Commands
11299 This filter supports the all above options as @ref{commands}.
11303 Deinterlace the input video ("estdif" stands for "Edge Slope
11304 Tracing Deinterlacing Filter").
11306 Spatial only filter that uses edge slope tracing algorithm
11307 to interpolate missing lines.
11308 It accepts the following parameters:
11312 The interlacing mode to adopt. It accepts one of the following values:
11316 Output one frame for each frame.
11318 Output one frame for each field.
11321 The default value is @code{field}.
11324 The picture field parity assumed for the input interlaced video. It accepts one
11325 of the following values:
11329 Assume the top field is first.
11331 Assume the bottom field is first.
11333 Enable automatic detection of field parity.
11336 The default value is @code{auto}.
11337 If the interlacing is unknown or the decoder does not export this information,
11338 top field first will be assumed.
11341 Specify which frames to deinterlace. Accepts one of the following
11346 Deinterlace all frames.
11348 Only deinterlace frames marked as interlaced.
11351 The default value is @code{all}.
11354 Specify the search radius for edge slope tracing. Default value is 1.
11355 Allowed range is from 1 to 15.
11358 Specify the search radius for best edge matching. Default value is 2.
11359 Allowed range is from 0 to 15.
11362 Specify the interpolation used. Default is 4-point interpolation. It accepts one
11363 of the following values:
11367 Two-point interpolation.
11369 Four-point interpolation.
11371 Six-point interpolation.
11375 @subsection Commands
11376 This filter supports same @ref{commands} as options.
11379 Adjust exposure of the video stream.
11381 The filter accepts the following options:
11385 Set the exposure correction in EV. Allowed range is from -3.0 to 3.0 EV
11386 Default value is 0 EV.
11389 Set the black level correction. Allowed range is from -1.0 to 1.0.
11390 Default value is 0.
11393 @subsection Commands
11395 This filter supports same @ref{commands} as options.
11397 @section extractplanes
11399 Extract color channel components from input video stream into
11400 separate grayscale video streams.
11402 The filter accepts the following option:
11406 Set plane(s) to extract.
11408 Available values for planes are:
11419 Choosing planes not available in the input will result in an error.
11420 That means you cannot select @code{r}, @code{g}, @code{b} planes
11421 with @code{y}, @code{u}, @code{v} planes at same time.
11424 @subsection Examples
11428 Extract luma, u and v color channel component from input video frame
11429 into 3 grayscale outputs:
11431 ffmpeg -i video.avi -filter_complex 'extractplanes=y+u+v[y][u][v]' -map '[y]' y.avi -map '[u]' u.avi -map '[v]' v.avi
11437 Apply a fade-in/out effect to the input video.
11439 It accepts the following parameters:
11443 The effect type can be either "in" for a fade-in, or "out" for a fade-out
11445 Default is @code{in}.
11447 @item start_frame, s
11448 Specify the number of the frame to start applying the fade
11449 effect at. Default is 0.
11452 The number of frames that the fade effect lasts. At the end of the
11453 fade-in effect, the output video will have the same intensity as the input video.
11454 At the end of the fade-out transition, the output video will be filled with the
11455 selected @option{color}.
11459 If set to 1, fade only alpha channel, if one exists on the input.
11460 Default value is 0.
11462 @item start_time, st
11463 Specify the timestamp (in seconds) of the frame to start to apply the fade
11464 effect. If both start_frame and start_time are specified, the fade will start at
11465 whichever comes last. Default is 0.
11468 The number of seconds for which the fade effect has to last. At the end of the
11469 fade-in effect the output video will have the same intensity as the input video,
11470 at the end of the fade-out transition the output video will be filled with the
11471 selected @option{color}.
11472 If both duration and nb_frames are specified, duration is used. Default is 0
11473 (nb_frames is used by default).
11476 Specify the color of the fade. Default is "black".
11479 @subsection Examples
11483 Fade in the first 30 frames of video:
11488 The command above is equivalent to:
11494 Fade out the last 45 frames of a 200-frame video:
11497 fade=type=out:start_frame=155:nb_frames=45
11501 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
11503 fade=in:0:25, fade=out:975:25
11507 Make the first 5 frames yellow, then fade in from frame 5-24:
11509 fade=in:5:20:color=yellow
11513 Fade in alpha over first 25 frames of video:
11515 fade=in:0:25:alpha=1
11519 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
11521 fade=t=in:st=5.5:d=0.5
11527 Denoise frames using 3D FFT (frequency domain filtering).
11529 The filter accepts the following options:
11533 Set the noise sigma constant. This sets denoising strength.
11534 Default value is 1. Allowed range is from 0 to 30.
11535 Using very high sigma with low overlap may give blocking artifacts.
11538 Set amount of denoising. By default all detected noise is reduced.
11539 Default value is 1. Allowed range is from 0 to 1.
11542 Set size of block, Default is 4, can be 3, 4, 5 or 6.
11543 Actual size of block in pixels is 2 to power of @var{block}, so by default
11544 block size in pixels is 2^4 which is 16.
11547 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
11550 Set number of previous frames to use for denoising. By default is set to 0.
11553 Set number of next frames to to use for denoising. By default is set to 0.
11556 Set planes which will be filtered, by default are all available filtered
11561 Apply arbitrary expressions to samples in frequency domain
11565 Adjust the dc value (gain) of the luma plane of the image. The filter
11566 accepts an integer value in range @code{0} to @code{1000}. The default
11567 value is set to @code{0}.
11570 Adjust the dc value (gain) of the 1st chroma plane of the image. The
11571 filter accepts an integer value in range @code{0} to @code{1000}. The
11572 default value is set to @code{0}.
11575 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
11576 filter accepts an integer value in range @code{0} to @code{1000}. The
11577 default value is set to @code{0}.
11580 Set the frequency domain weight expression for the luma plane.
11583 Set the frequency domain weight expression for the 1st chroma plane.
11586 Set the frequency domain weight expression for the 2nd chroma plane.
11589 Set when the expressions are evaluated.
11591 It accepts the following values:
11594 Only evaluate expressions once during the filter initialization.
11597 Evaluate expressions for each incoming frame.
11600 Default value is @samp{init}.
11602 The filter accepts the following variables:
11605 The coordinates of the current sample.
11609 The width and height of the image.
11612 The number of input frame, starting from 0.
11615 @subsection Examples
11621 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
11627 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
11633 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
11639 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
11646 Extract a single field from an interlaced image using stride
11647 arithmetic to avoid wasting CPU time. The output frames are marked as
11650 The filter accepts the following options:
11654 Specify whether to extract the top (if the value is @code{0} or
11655 @code{top}) or the bottom field (if the value is @code{1} or
11661 Create new frames by copying the top and bottom fields from surrounding frames
11662 supplied as numbers by the hint file.
11666 Set file containing hints: absolute/relative frame numbers.
11668 There must be one line for each frame in a clip. Each line must contain two
11669 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
11670 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
11671 is current frame number for @code{absolute} mode or out of [-1, 1] range
11672 for @code{relative} mode. First number tells from which frame to pick up top
11673 field and second number tells from which frame to pick up bottom field.
11675 If optionally followed by @code{+} output frame will be marked as interlaced,
11676 else if followed by @code{-} output frame will be marked as progressive, else
11677 it will be marked same as input frame.
11678 If optionally followed by @code{t} output frame will use only top field, or in
11679 case of @code{b} it will use only bottom field.
11680 If line starts with @code{#} or @code{;} that line is skipped.
11683 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
11686 Example of first several lines of @code{hint} file for @code{relative} mode:
11688 0,0 - # first frame
11689 1,0 - # second frame, use third's frame top field and second's frame bottom field
11690 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
11705 @section fieldmatch
11707 Field matching filter for inverse telecine. It is meant to reconstruct the
11708 progressive frames from a telecined stream. The filter does not drop duplicated
11709 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
11710 followed by a decimation filter such as @ref{decimate} in the filtergraph.
11712 The separation of the field matching and the decimation is notably motivated by
11713 the possibility of inserting a de-interlacing filter fallback between the two.
11714 If the source has mixed telecined and real interlaced content,
11715 @code{fieldmatch} will not be able to match fields for the interlaced parts.
11716 But these remaining combed frames will be marked as interlaced, and thus can be
11717 de-interlaced by a later filter such as @ref{yadif} before decimation.
11719 In addition to the various configuration options, @code{fieldmatch} can take an
11720 optional second stream, activated through the @option{ppsrc} option. If
11721 enabled, the frames reconstruction will be based on the fields and frames from
11722 this second stream. This allows the first input to be pre-processed in order to
11723 help the various algorithms of the filter, while keeping the output lossless
11724 (assuming the fields are matched properly). Typically, a field-aware denoiser,
11725 or brightness/contrast adjustments can help.
11727 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
11728 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
11729 which @code{fieldmatch} is based on. While the semantic and usage are very
11730 close, some behaviour and options names can differ.
11732 The @ref{decimate} filter currently only works for constant frame rate input.
11733 If your input has mixed telecined (30fps) and progressive content with a lower
11734 framerate like 24fps use the following filterchain to produce the necessary cfr
11735 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
11737 The filter accepts the following options:
11741 Specify the assumed field order of the input stream. Available values are:
11745 Auto detect parity (use FFmpeg's internal parity value).
11747 Assume bottom field first.
11749 Assume top field first.
11752 Note that it is sometimes recommended not to trust the parity announced by the
11755 Default value is @var{auto}.
11758 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
11759 sense that it won't risk creating jerkiness due to duplicate frames when
11760 possible, but if there are bad edits or blended fields it will end up
11761 outputting combed frames when a good match might actually exist. On the other
11762 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
11763 but will almost always find a good frame if there is one. The other values are
11764 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
11765 jerkiness and creating duplicate frames versus finding good matches in sections
11766 with bad edits, orphaned fields, blended fields, etc.
11768 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
11770 Available values are:
11774 2-way matching (p/c)
11776 2-way matching, and trying 3rd match if still combed (p/c + n)
11778 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
11780 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
11781 still combed (p/c + n + u/b)
11783 3-way matching (p/c/n)
11785 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
11786 detected as combed (p/c/n + u/b)
11789 The parenthesis at the end indicate the matches that would be used for that
11790 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
11793 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
11796 Default value is @var{pc_n}.
11799 Mark the main input stream as a pre-processed input, and enable the secondary
11800 input stream as the clean source to pick the fields from. See the filter
11801 introduction for more details. It is similar to the @option{clip2} feature from
11804 Default value is @code{0} (disabled).
11807 Set the field to match from. It is recommended to set this to the same value as
11808 @option{order} unless you experience matching failures with that setting. In
11809 certain circumstances changing the field that is used to match from can have a
11810 large impact on matching performance. Available values are:
11814 Automatic (same value as @option{order}).
11816 Match from the bottom field.
11818 Match from the top field.
11821 Default value is @var{auto}.
11824 Set whether or not chroma is included during the match comparisons. In most
11825 cases it is recommended to leave this enabled. You should set this to @code{0}
11826 only if your clip has bad chroma problems such as heavy rainbowing or other
11827 artifacts. Setting this to @code{0} could also be used to speed things up at
11828 the cost of some accuracy.
11830 Default value is @code{1}.
11834 These define an exclusion band which excludes the lines between @option{y0} and
11835 @option{y1} from being included in the field matching decision. An exclusion
11836 band can be used to ignore subtitles, a logo, or other things that may
11837 interfere with the matching. @option{y0} sets the starting scan line and
11838 @option{y1} sets the ending line; all lines in between @option{y0} and
11839 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
11840 @option{y0} and @option{y1} to the same value will disable the feature.
11841 @option{y0} and @option{y1} defaults to @code{0}.
11844 Set the scene change detection threshold as a percentage of maximum change on
11845 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
11846 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
11847 @option{scthresh} is @code{[0.0, 100.0]}.
11849 Default value is @code{12.0}.
11852 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
11853 account the combed scores of matches when deciding what match to use as the
11854 final match. Available values are:
11858 No final matching based on combed scores.
11860 Combed scores are only used when a scene change is detected.
11862 Use combed scores all the time.
11865 Default is @var{sc}.
11868 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
11869 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
11870 Available values are:
11874 No forced calculation.
11876 Force p/c/n calculations.
11878 Force p/c/n/u/b calculations.
11881 Default value is @var{none}.
11884 This is the area combing threshold used for combed frame detection. This
11885 essentially controls how "strong" or "visible" combing must be to be detected.
11886 Larger values mean combing must be more visible and smaller values mean combing
11887 can be less visible or strong and still be detected. Valid settings are from
11888 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
11889 be detected as combed). This is basically a pixel difference value. A good
11890 range is @code{[8, 12]}.
11892 Default value is @code{9}.
11895 Sets whether or not chroma is considered in the combed frame decision. Only
11896 disable this if your source has chroma problems (rainbowing, etc.) that are
11897 causing problems for the combed frame detection with chroma enabled. Actually,
11898 using @option{chroma}=@var{0} is usually more reliable, except for the case
11899 where there is chroma only combing in the source.
11901 Default value is @code{0}.
11905 Respectively set the x-axis and y-axis size of the window used during combed
11906 frame detection. This has to do with the size of the area in which
11907 @option{combpel} pixels are required to be detected as combed for a frame to be
11908 declared combed. See the @option{combpel} parameter description for more info.
11909 Possible values are any number that is a power of 2 starting at 4 and going up
11912 Default value is @code{16}.
11915 The number of combed pixels inside any of the @option{blocky} by
11916 @option{blockx} size blocks on the frame for the frame to be detected as
11917 combed. While @option{cthresh} controls how "visible" the combing must be, this
11918 setting controls "how much" combing there must be in any localized area (a
11919 window defined by the @option{blockx} and @option{blocky} settings) on the
11920 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
11921 which point no frames will ever be detected as combed). This setting is known
11922 as @option{MI} in TFM/VFM vocabulary.
11924 Default value is @code{80}.
11927 @anchor{p/c/n/u/b meaning}
11928 @subsection p/c/n/u/b meaning
11930 @subsubsection p/c/n
11932 We assume the following telecined stream:
11935 Top fields: 1 2 2 3 4
11936 Bottom fields: 1 2 3 4 4
11939 The numbers correspond to the progressive frame the fields relate to. Here, the
11940 first two frames are progressive, the 3rd and 4th are combed, and so on.
11942 When @code{fieldmatch} is configured to run a matching from bottom
11943 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11948 B 1 2 3 4 4 <-- matching reference
11957 As a result of the field matching, we can see that some frames get duplicated.
11958 To perform a complete inverse telecine, you need to rely on a decimation filter
11959 after this operation. See for instance the @ref{decimate} filter.
11961 The same operation now matching from top fields (@option{field}=@var{top})
11966 T 1 2 2 3 4 <-- matching reference
11976 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11977 basically, they refer to the frame and field of the opposite parity:
11980 @item @var{p} matches the field of the opposite parity in the previous frame
11981 @item @var{c} matches the field of the opposite parity in the current frame
11982 @item @var{n} matches the field of the opposite parity in the next frame
11987 The @var{u} and @var{b} matching are a bit special in the sense that they match
11988 from the opposite parity flag. In the following examples, we assume that we are
11989 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11990 'x' is placed above and below each matched fields.
11992 With bottom matching (@option{field}=@var{bottom}):
11997 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11998 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
12006 With top matching (@option{field}=@var{top}):
12011 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
12012 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
12020 @subsection Examples
12022 Simple IVTC of a top field first telecined stream:
12024 fieldmatch=order=tff:combmatch=none, decimate
12027 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
12029 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
12032 @section fieldorder
12034 Transform the field order of the input video.
12036 It accepts the following parameters:
12041 The output field order. Valid values are @var{tff} for top field first or @var{bff}
12042 for bottom field first.
12045 The default value is @samp{tff}.
12047 The transformation is done by shifting the picture content up or down
12048 by one line, and filling the remaining line with appropriate picture content.
12049 This method is consistent with most broadcast field order converters.
12051 If the input video is not flagged as being interlaced, or it is already
12052 flagged as being of the required output field order, then this filter does
12053 not alter the incoming video.
12055 It is very useful when converting to or from PAL DV material,
12056 which is bottom field first.
12060 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
12063 @section fifo, afifo
12065 Buffer input images and send them when they are requested.
12067 It is mainly useful when auto-inserted by the libavfilter
12070 It does not take parameters.
12072 @section fillborders
12074 Fill borders of the input video, without changing video stream dimensions.
12075 Sometimes video can have garbage at the four edges and you may not want to
12076 crop video input to keep size multiple of some number.
12078 This filter accepts the following options:
12082 Number of pixels to fill from left border.
12085 Number of pixels to fill from right border.
12088 Number of pixels to fill from top border.
12091 Number of pixels to fill from bottom border.
12096 It accepts the following values:
12099 fill pixels using outermost pixels
12102 fill pixels using mirroring (half sample symmetric)
12105 fill pixels with constant value
12108 fill pixels using reflecting (whole sample symmetric)
12111 fill pixels using wrapping
12114 fade pixels to constant value
12117 Default is @var{smear}.
12120 Set color for pixels in fixed or fade mode. Default is @var{black}.
12123 @subsection Commands
12124 This filter supports same @ref{commands} as options.
12125 The command accepts the same syntax of the corresponding option.
12127 If the specified expression is not valid, it is kept at its current
12132 Find a rectangular object
12134 It accepts the following options:
12138 Filepath of the object image, needs to be in gray8.
12141 Detection threshold, default is 0.5.
12144 Number of mipmaps, default is 3.
12146 @item xmin, ymin, xmax, ymax
12147 Specifies the rectangle in which to search.
12150 @subsection Examples
12154 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
12156 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
12162 Flood area with values of same pixel components with another values.
12164 It accepts the following options:
12167 Set pixel x coordinate.
12170 Set pixel y coordinate.
12173 Set source #0 component value.
12176 Set source #1 component value.
12179 Set source #2 component value.
12182 Set source #3 component value.
12185 Set destination #0 component value.
12188 Set destination #1 component value.
12191 Set destination #2 component value.
12194 Set destination #3 component value.
12200 Convert the input video to one of the specified pixel formats.
12201 Libavfilter will try to pick one that is suitable as input to
12204 It accepts the following parameters:
12208 A '|'-separated list of pixel format names, such as
12209 "pix_fmts=yuv420p|monow|rgb24".
12213 @subsection Examples
12217 Convert the input video to the @var{yuv420p} format
12219 format=pix_fmts=yuv420p
12222 Convert the input video to any of the formats in the list
12224 format=pix_fmts=yuv420p|yuv444p|yuv410p
12231 Convert the video to specified constant frame rate by duplicating or dropping
12232 frames as necessary.
12234 It accepts the following parameters:
12238 The desired output frame rate. The default is @code{25}.
12241 Assume the first PTS should be the given value, in seconds. This allows for
12242 padding/trimming at the start of stream. By default, no assumption is made
12243 about the first frame's expected PTS, so no padding or trimming is done.
12244 For example, this could be set to 0 to pad the beginning with duplicates of
12245 the first frame if a video stream starts after the audio stream or to trim any
12246 frames with a negative PTS.
12249 Timestamp (PTS) rounding method.
12251 Possible values are:
12258 round towards -infinity
12260 round towards +infinity
12264 The default is @code{near}.
12267 Action performed when reading the last frame.
12269 Possible values are:
12272 Use same timestamp rounding method as used for other frames.
12274 Pass through last frame if input duration has not been reached yet.
12276 The default is @code{round}.
12280 Alternatively, the options can be specified as a flat string:
12281 @var{fps}[:@var{start_time}[:@var{round}]].
12283 See also the @ref{setpts} filter.
12285 @subsection Examples
12289 A typical usage in order to set the fps to 25:
12295 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
12297 fps=fps=film:round=near
12303 Pack two different video streams into a stereoscopic video, setting proper
12304 metadata on supported codecs. The two views should have the same size and
12305 framerate and processing will stop when the shorter video ends. Please note
12306 that you may conveniently adjust view properties with the @ref{scale} and
12309 It accepts the following parameters:
12313 The desired packing format. Supported values are:
12318 The views are next to each other (default).
12321 The views are on top of each other.
12324 The views are packed by line.
12327 The views are packed by column.
12330 The views are temporally interleaved.
12339 # Convert left and right views into a frame-sequential video
12340 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
12342 # Convert views into a side-by-side video with the same output resolution as the input
12343 ffmpeg -i LEFT -i RIGHT -filter_complex [0:v]scale=w=iw/2[left],[1:v]scale=w=iw/2[right],[left][right]framepack=sbs OUTPUT
12348 Change the frame rate by interpolating new video output frames from the source
12351 This filter is not designed to function correctly with interlaced media. If
12352 you wish to change the frame rate of interlaced media then you are required
12353 to deinterlace before this filter and re-interlace after this filter.
12355 A description of the accepted options follows.
12359 Specify the output frames per second. This option can also be specified
12360 as a value alone. The default is @code{50}.
12363 Specify the start of a range where the output frame will be created as a
12364 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12365 the default is @code{15}.
12368 Specify the end of a range where the output frame will be created as a
12369 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12370 the default is @code{240}.
12373 Specify the level at which a scene change is detected as a value between
12374 0 and 100 to indicate a new scene; a low value reflects a low
12375 probability for the current frame to introduce a new scene, while a higher
12376 value means the current frame is more likely to be one.
12377 The default is @code{8.2}.
12380 Specify flags influencing the filter process.
12382 Available value for @var{flags} is:
12385 @item scene_change_detect, scd
12386 Enable scene change detection using the value of the option @var{scene}.
12387 This flag is enabled by default.
12393 Select one frame every N-th frame.
12395 This filter accepts the following option:
12398 Select frame after every @code{step} frames.
12399 Allowed values are positive integers higher than 0. Default value is @code{1}.
12402 @section freezedetect
12404 Detect frozen video.
12406 This filter logs a message and sets frame metadata when it detects that the
12407 input video has no significant change in content during a specified duration.
12408 Video freeze detection calculates the mean average absolute difference of all
12409 the components of video frames and compares it to a noise floor.
12411 The printed times and duration are expressed in seconds. The
12412 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
12413 whose timestamp equals or exceeds the detection duration and it contains the
12414 timestamp of the first frame of the freeze. The
12415 @code{lavfi.freezedetect.freeze_duration} and
12416 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
12419 The filter accepts the following options:
12423 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
12424 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
12428 Set freeze duration until notification (default is 2 seconds).
12431 @section freezeframes
12433 Freeze video frames.
12435 This filter freezes video frames using frame from 2nd input.
12437 The filter accepts the following options:
12441 Set number of first frame from which to start freeze.
12444 Set number of last frame from which to end freeze.
12447 Set number of frame from 2nd input which will be used instead of replaced frames.
12453 Apply a frei0r effect to the input video.
12455 To enable the compilation of this filter, you need to install the frei0r
12456 header and configure FFmpeg with @code{--enable-frei0r}.
12458 It accepts the following parameters:
12463 The name of the frei0r effect to load. If the environment variable
12464 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
12465 directories specified by the colon-separated list in @env{FREI0R_PATH}.
12466 Otherwise, the standard frei0r paths are searched, in this order:
12467 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
12468 @file{/usr/lib/frei0r-1/}.
12470 @item filter_params
12471 A '|'-separated list of parameters to pass to the frei0r effect.
12475 A frei0r effect parameter can be a boolean (its value is either
12476 "y" or "n"), a double, a color (specified as
12477 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
12478 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
12479 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
12480 a position (specified as @var{X}/@var{Y}, where
12481 @var{X} and @var{Y} are floating point numbers) and/or a string.
12483 The number and types of parameters depend on the loaded effect. If an
12484 effect parameter is not specified, the default value is set.
12486 @subsection Examples
12490 Apply the distort0r effect, setting the first two double parameters:
12492 frei0r=filter_name=distort0r:filter_params=0.5|0.01
12496 Apply the colordistance effect, taking a color as the first parameter:
12498 frei0r=colordistance:0.2/0.3/0.4
12499 frei0r=colordistance:violet
12500 frei0r=colordistance:0x112233
12504 Apply the perspective effect, specifying the top left and top right image
12507 frei0r=perspective:0.2/0.2|0.8/0.2
12511 For more information, see
12512 @url{http://frei0r.dyne.org}
12514 @subsection Commands
12516 This filter supports the @option{filter_params} option as @ref{commands}.
12520 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
12522 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
12523 processing filter, one of them is performed once per block, not per pixel.
12524 This allows for much higher speed.
12526 The filter accepts the following options:
12530 Set quality. This option defines the number of levels for averaging. It accepts
12531 an integer in the range 4-5. Default value is @code{4}.
12534 Force a constant quantization parameter. It accepts an integer in range 0-63.
12535 If not set, the filter will use the QP from the video stream (if available).
12538 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
12539 more details but also more artifacts, while higher values make the image smoother
12540 but also blurrier. Default value is @code{0} − PSNR optimal.
12542 @item use_bframe_qp
12543 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12544 option may cause flicker since the B-Frames have often larger QP. Default is
12545 @code{0} (not enabled).
12551 Apply Gaussian blur filter.
12553 The filter accepts the following options:
12557 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
12560 Set number of steps for Gaussian approximation. Default is @code{1}.
12563 Set which planes to filter. By default all planes are filtered.
12566 Set vertical sigma, if negative it will be same as @code{sigma}.
12567 Default is @code{-1}.
12570 @subsection Commands
12571 This filter supports same commands as options.
12572 The command accepts the same syntax of the corresponding option.
12574 If the specified expression is not valid, it is kept at its current
12579 Apply generic equation to each pixel.
12581 The filter accepts the following options:
12584 @item lum_expr, lum
12585 Set the luminance expression.
12587 Set the chrominance blue expression.
12589 Set the chrominance red expression.
12590 @item alpha_expr, a
12591 Set the alpha expression.
12593 Set the red expression.
12594 @item green_expr, g
12595 Set the green expression.
12597 Set the blue expression.
12600 The colorspace is selected according to the specified options. If one
12601 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
12602 options is specified, the filter will automatically select a YCbCr
12603 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
12604 @option{blue_expr} options is specified, it will select an RGB
12607 If one of the chrominance expression is not defined, it falls back on the other
12608 one. If no alpha expression is specified it will evaluate to opaque value.
12609 If none of chrominance expressions are specified, they will evaluate
12610 to the luminance expression.
12612 The expressions can use the following variables and functions:
12616 The sequential number of the filtered frame, starting from @code{0}.
12620 The coordinates of the current sample.
12624 The width and height of the image.
12628 Width and height scale depending on the currently filtered plane. It is the
12629 ratio between the corresponding luma plane number of pixels and the current
12630 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
12631 @code{0.5,0.5} for chroma planes.
12634 Time of the current frame, expressed in seconds.
12637 Return the value of the pixel at location (@var{x},@var{y}) of the current
12641 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
12645 Return the value of the pixel at location (@var{x},@var{y}) of the
12646 blue-difference chroma plane. Return 0 if there is no such plane.
12649 Return the value of the pixel at location (@var{x},@var{y}) of the
12650 red-difference chroma plane. Return 0 if there is no such plane.
12655 Return the value of the pixel at location (@var{x},@var{y}) of the
12656 red/green/blue component. Return 0 if there is no such component.
12659 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
12660 plane. Return 0 if there is no such plane.
12662 @item psum(x,y), lumsum(x, y), cbsum(x,y), crsum(x,y), rsum(x,y), gsum(x,y), bsum(x,y), alphasum(x,y)
12663 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
12664 sums of samples within a rectangle. See the functions without the sum postfix.
12666 @item interpolation
12667 Set one of interpolation methods:
12672 Default is bilinear.
12675 For functions, if @var{x} and @var{y} are outside the area, the value will be
12676 automatically clipped to the closer edge.
12678 Please note that this filter can use multiple threads in which case each slice
12679 will have its own expression state. If you want to use only a single expression
12680 state because your expressions depend on previous state then you should limit
12681 the number of filter threads to 1.
12683 @subsection Examples
12687 Flip the image horizontally:
12693 Generate a bidimensional sine wave, with angle @code{PI/3} and a
12694 wavelength of 100 pixels:
12696 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
12700 Generate a fancy enigmatic moving light:
12702 nullsrc=s=256x256,geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128
12706 Generate a quick emboss effect:
12708 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
12712 Modify RGB components depending on pixel position:
12714 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
12718 Create a radial gradient that is the same size as the input (also see
12719 the @ref{vignette} filter):
12721 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
12727 Fix the banding artifacts that are sometimes introduced into nearly flat
12728 regions by truncation to 8-bit color depth.
12729 Interpolate the gradients that should go where the bands are, and
12732 It is designed for playback only. Do not use it prior to
12733 lossy compression, because compression tends to lose the dither and
12734 bring back the bands.
12736 It accepts the following parameters:
12741 The maximum amount by which the filter will change any one pixel. This is also
12742 the threshold for detecting nearly flat regions. Acceptable values range from
12743 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
12747 The neighborhood to fit the gradient to. A larger radius makes for smoother
12748 gradients, but also prevents the filter from modifying the pixels near detailed
12749 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
12750 values will be clipped to the valid range.
12754 Alternatively, the options can be specified as a flat string:
12755 @var{strength}[:@var{radius}]
12757 @subsection Examples
12761 Apply the filter with a @code{3.5} strength and radius of @code{8}:
12767 Specify radius, omitting the strength (which will fall-back to the default
12775 @anchor{graphmonitor}
12776 @section graphmonitor
12777 Show various filtergraph stats.
12779 With this filter one can debug complete filtergraph.
12780 Especially issues with links filling with queued frames.
12782 The filter accepts the following options:
12786 Set video output size. Default is @var{hd720}.
12789 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
12792 Set output mode, can be @var{fulll} or @var{compact}.
12793 In @var{compact} mode only filters with some queued frames have displayed stats.
12796 Set flags which enable which stats are shown in video.
12798 Available values for flags are:
12801 Display number of queued frames in each link.
12803 @item frame_count_in
12804 Display number of frames taken from filter.
12806 @item frame_count_out
12807 Display number of frames given out from filter.
12810 Display current filtered frame pts.
12813 Display current filtered frame time.
12816 Display time base for filter link.
12819 Display used format for filter link.
12822 Display video size or number of audio channels in case of audio used by filter link.
12825 Display video frame rate or sample rate in case of audio used by filter link.
12828 Display link output status.
12832 Set upper limit for video rate of output stream, Default value is @var{25}.
12833 This guarantee that output video frame rate will not be higher than this value.
12837 A color constancy variation filter which estimates scene illumination via grey edge algorithm
12838 and corrects the scene colors accordingly.
12840 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
12842 The filter accepts the following options:
12846 The order of differentiation to be applied on the scene. Must be chosen in the range
12847 [0,2] and default value is 1.
12850 The Minkowski parameter to be used for calculating the Minkowski distance. Must
12851 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
12852 max value instead of calculating Minkowski distance.
12855 The standard deviation of Gaussian blur to be applied on the scene. Must be
12856 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
12857 can't be equal to 0 if @var{difford} is greater than 0.
12860 @subsection Examples
12866 greyedge=difford=1:minknorm=5:sigma=2
12872 greyedge=difford=1:minknorm=0:sigma=2
12880 Apply a Hald CLUT to a video stream.
12882 First input is the video stream to process, and second one is the Hald CLUT.
12883 The Hald CLUT input can be a simple picture or a complete video stream.
12885 The filter accepts the following options:
12889 Force termination when the shortest input terminates. Default is @code{0}.
12891 Continue applying the last CLUT after the end of the stream. A value of
12892 @code{0} disable the filter after the last frame of the CLUT is reached.
12893 Default is @code{1}.
12896 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
12897 filters share the same internals).
12899 This filter also supports the @ref{framesync} options.
12901 More information about the Hald CLUT can be found on Eskil Steenberg's website
12902 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
12904 @subsection Commands
12906 This filter supports the @code{interp} option as @ref{commands}.
12908 @subsection Workflow examples
12910 @subsubsection Hald CLUT video stream
12912 Generate an identity Hald CLUT stream altered with various effects:
12914 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "hue=H=2*PI*t:s=sin(2*PI*t)+1, curves=cross_process" -t 10 -c:v ffv1 clut.nut
12917 Note: make sure you use a lossless codec.
12919 Then use it with @code{haldclut} to apply it on some random stream:
12921 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
12924 The Hald CLUT will be applied to the 10 first seconds (duration of
12925 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
12926 to the remaining frames of the @code{mandelbrot} stream.
12928 @subsubsection Hald CLUT with preview
12930 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
12931 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
12932 biggest possible square starting at the top left of the picture. The remaining
12933 padding pixels (bottom or right) will be ignored. This area can be used to add
12934 a preview of the Hald CLUT.
12936 Typically, the following generated Hald CLUT will be supported by the
12937 @code{haldclut} filter:
12940 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
12941 pad=iw+320 [padded_clut];
12942 smptebars=s=320x256, split [a][b];
12943 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
12944 [main][b] overlay=W-320" -frames:v 1 clut.png
12947 It contains the original and a preview of the effect of the CLUT: SMPTE color
12948 bars are displayed on the right-top, and below the same color bars processed by
12951 Then, the effect of this Hald CLUT can be visualized with:
12953 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
12958 Flip the input video horizontally.
12960 For example, to horizontally flip the input video with @command{ffmpeg}:
12962 ffmpeg -i in.avi -vf "hflip" out.avi
12966 This filter applies a global color histogram equalization on a
12969 It can be used to correct video that has a compressed range of pixel
12970 intensities. The filter redistributes the pixel intensities to
12971 equalize their distribution across the intensity range. It may be
12972 viewed as an "automatically adjusting contrast filter". This filter is
12973 useful only for correcting degraded or poorly captured source
12976 The filter accepts the following options:
12980 Determine the amount of equalization to be applied. As the strength
12981 is reduced, the distribution of pixel intensities more-and-more
12982 approaches that of the input frame. The value must be a float number
12983 in the range [0,1] and defaults to 0.200.
12986 Set the maximum intensity that can generated and scale the output
12987 values appropriately. The strength should be set as desired and then
12988 the intensity can be limited if needed to avoid washing-out. The value
12989 must be a float number in the range [0,1] and defaults to 0.210.
12992 Set the antibanding level. If enabled the filter will randomly vary
12993 the luminance of output pixels by a small amount to avoid banding of
12994 the histogram. Possible values are @code{none}, @code{weak} or
12995 @code{strong}. It defaults to @code{none}.
13001 Compute and draw a color distribution histogram for the input video.
13003 The computed histogram is a representation of the color component
13004 distribution in an image.
13006 Standard histogram displays the color components distribution in an image.
13007 Displays color graph for each color component. Shows distribution of
13008 the Y, U, V, A or R, G, B components, depending on input format, in the
13009 current frame. Below each graph a color component scale meter is shown.
13011 The filter accepts the following options:
13015 Set height of level. Default value is @code{200}.
13016 Allowed range is [50, 2048].
13019 Set height of color scale. Default value is @code{12}.
13020 Allowed range is [0, 40].
13024 It accepts the following values:
13027 Per color component graphs are placed below each other.
13030 Per color component graphs are placed side by side.
13033 Presents information identical to that in the @code{parade}, except
13034 that the graphs representing color components are superimposed directly
13037 Default is @code{stack}.
13040 Set mode. Can be either @code{linear}, or @code{logarithmic}.
13041 Default is @code{linear}.
13044 Set what color components to display.
13045 Default is @code{7}.
13048 Set foreground opacity. Default is @code{0.7}.
13051 Set background opacity. Default is @code{0.5}.
13054 @subsection Examples
13059 Calculate and draw histogram:
13061 ffplay -i input -vf histogram
13069 This is a high precision/quality 3d denoise filter. It aims to reduce
13070 image noise, producing smooth images and making still images really
13071 still. It should enhance compressibility.
13073 It accepts the following optional parameters:
13077 A non-negative floating point number which specifies spatial luma strength.
13078 It defaults to 4.0.
13080 @item chroma_spatial
13081 A non-negative floating point number which specifies spatial chroma strength.
13082 It defaults to 3.0*@var{luma_spatial}/4.0.
13085 A floating point number which specifies luma temporal strength. It defaults to
13086 6.0*@var{luma_spatial}/4.0.
13089 A floating point number which specifies chroma temporal strength. It defaults to
13090 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
13093 @subsection Commands
13094 This filter supports same @ref{commands} as options.
13095 The command accepts the same syntax of the corresponding option.
13097 If the specified expression is not valid, it is kept at its current
13100 @anchor{hwdownload}
13101 @section hwdownload
13103 Download hardware frames to system memory.
13105 The input must be in hardware frames, and the output a non-hardware format.
13106 Not all formats will be supported on the output - it may be necessary to insert
13107 an additional @option{format} filter immediately following in the graph to get
13108 the output in a supported format.
13112 Map hardware frames to system memory or to another device.
13114 This filter has several different modes of operation; which one is used depends
13115 on the input and output formats:
13118 Hardware frame input, normal frame output
13120 Map the input frames to system memory and pass them to the output. If the
13121 original hardware frame is later required (for example, after overlaying
13122 something else on part of it), the @option{hwmap} filter can be used again
13123 in the next mode to retrieve it.
13125 Normal frame input, hardware frame output
13127 If the input is actually a software-mapped hardware frame, then unmap it -
13128 that is, return the original hardware frame.
13130 Otherwise, a device must be provided. Create new hardware surfaces on that
13131 device for the output, then map them back to the software format at the input
13132 and give those frames to the preceding filter. This will then act like the
13133 @option{hwupload} filter, but may be able to avoid an additional copy when
13134 the input is already in a compatible format.
13136 Hardware frame input and output
13138 A device must be supplied for the output, either directly or with the
13139 @option{derive_device} option. The input and output devices must be of
13140 different types and compatible - the exact meaning of this is
13141 system-dependent, but typically it means that they must refer to the same
13142 underlying hardware context (for example, refer to the same graphics card).
13144 If the input frames were originally created on the output device, then unmap
13145 to retrieve the original frames.
13147 Otherwise, map the frames to the output device - create new hardware frames
13148 on the output corresponding to the frames on the input.
13151 The following additional parameters are accepted:
13155 Set the frame mapping mode. Some combination of:
13158 The mapped frame should be readable.
13160 The mapped frame should be writeable.
13162 The mapping will always overwrite the entire frame.
13164 This may improve performance in some cases, as the original contents of the
13165 frame need not be loaded.
13167 The mapping must not involve any copying.
13169 Indirect mappings to copies of frames are created in some cases where either
13170 direct mapping is not possible or it would have unexpected properties.
13171 Setting this flag ensures that the mapping is direct and will fail if that is
13174 Defaults to @var{read+write} if not specified.
13176 @item derive_device @var{type}
13177 Rather than using the device supplied at initialisation, instead derive a new
13178 device of type @var{type} from the device the input frames exist on.
13181 In a hardware to hardware mapping, map in reverse - create frames in the sink
13182 and map them back to the source. This may be necessary in some cases where
13183 a mapping in one direction is required but only the opposite direction is
13184 supported by the devices being used.
13186 This option is dangerous - it may break the preceding filter in undefined
13187 ways if there are any additional constraints on that filter's output.
13188 Do not use it without fully understanding the implications of its use.
13194 Upload system memory frames to hardware surfaces.
13196 The device to upload to must be supplied when the filter is initialised. If
13197 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
13198 option or with the @option{derive_device} option. The input and output devices
13199 must be of different types and compatible - the exact meaning of this is
13200 system-dependent, but typically it means that they must refer to the same
13201 underlying hardware context (for example, refer to the same graphics card).
13203 The following additional parameters are accepted:
13206 @item derive_device @var{type}
13207 Rather than using the device supplied at initialisation, instead derive a new
13208 device of type @var{type} from the device the input frames exist on.
13211 @anchor{hwupload_cuda}
13212 @section hwupload_cuda
13214 Upload system memory frames to a CUDA device.
13216 It accepts the following optional parameters:
13220 The number of the CUDA device to use
13225 Apply a high-quality magnification filter designed for pixel art. This filter
13226 was originally created by Maxim Stepin.
13228 It accepts the following option:
13232 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
13233 @code{hq3x} and @code{4} for @code{hq4x}.
13234 Default is @code{3}.
13238 Stack input videos horizontally.
13240 All streams must be of same pixel format and of same height.
13242 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13243 to create same output.
13245 The filter accepts the following option:
13249 Set number of input streams. Default is 2.
13252 If set to 1, force the output to terminate when the shortest input
13253 terminates. Default value is 0.
13258 Modify the hue and/or the saturation of the input.
13260 It accepts the following parameters:
13264 Specify the hue angle as a number of degrees. It accepts an expression,
13265 and defaults to "0".
13268 Specify the saturation in the [-10,10] range. It accepts an expression and
13272 Specify the hue angle as a number of radians. It accepts an
13273 expression, and defaults to "0".
13276 Specify the brightness in the [-10,10] range. It accepts an expression and
13280 @option{h} and @option{H} are mutually exclusive, and can't be
13281 specified at the same time.
13283 The @option{b}, @option{h}, @option{H} and @option{s} option values are
13284 expressions containing the following constants:
13288 frame count of the input frame starting from 0
13291 presentation timestamp of the input frame expressed in time base units
13294 frame rate of the input video, NAN if the input frame rate is unknown
13297 timestamp expressed in seconds, NAN if the input timestamp is unknown
13300 time base of the input video
13303 @subsection Examples
13307 Set the hue to 90 degrees and the saturation to 1.0:
13313 Same command but expressing the hue in radians:
13319 Rotate hue and make the saturation swing between 0
13320 and 2 over a period of 1 second:
13322 hue="H=2*PI*t: s=sin(2*PI*t)+1"
13326 Apply a 3 seconds saturation fade-in effect starting at 0:
13328 hue="s=min(t/3\,1)"
13331 The general fade-in expression can be written as:
13333 hue="s=min(0\, max((t-START)/DURATION\, 1))"
13337 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
13339 hue="s=max(0\, min(1\, (8-t)/3))"
13342 The general fade-out expression can be written as:
13344 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
13349 @subsection Commands
13351 This filter supports the following commands:
13357 Modify the hue and/or the saturation and/or brightness of the input video.
13358 The command accepts the same syntax of the corresponding option.
13360 If the specified expression is not valid, it is kept at its current
13364 @section hysteresis
13366 Grow first stream into second stream by connecting components.
13367 This makes it possible to build more robust edge masks.
13369 This filter accepts the following options:
13373 Set which planes will be processed as bitmap, unprocessed planes will be
13374 copied from first stream.
13375 By default value 0xf, all planes will be processed.
13378 Set threshold which is used in filtering. If pixel component value is higher than
13379 this value filter algorithm for connecting components is activated.
13380 By default value is 0.
13383 The @code{hysteresis} filter also supports the @ref{framesync} options.
13387 Obtain the identity score between two input videos.
13389 This filter takes two input videos.
13391 Both input videos must have the same resolution and pixel format for
13392 this filter to work correctly. Also it assumes that both inputs
13393 have the same number of frames, which are compared one by one.
13395 The obtained per component, average, min and max identity score is printed through
13396 the logging system.
13398 The filter stores the calculated identity scores of each frame in frame metadata.
13400 In the below example the input file @file{main.mpg} being processed is compared
13401 with the reference file @file{ref.mpg}.
13404 ffmpeg -i main.mpg -i ref.mpg -lavfi identity -f null -
13409 Detect video interlacing type.
13411 This filter tries to detect if the input frames are interlaced, progressive,
13412 top or bottom field first. It will also try to detect fields that are
13413 repeated between adjacent frames (a sign of telecine).
13415 Single frame detection considers only immediately adjacent frames when classifying each frame.
13416 Multiple frame detection incorporates the classification history of previous frames.
13418 The filter will log these metadata values:
13421 @item single.current_frame
13422 Detected type of current frame using single-frame detection. One of:
13423 ``tff'' (top field first), ``bff'' (bottom field first),
13424 ``progressive'', or ``undetermined''
13427 Cumulative number of frames detected as top field first using single-frame detection.
13430 Cumulative number of frames detected as top field first using multiple-frame detection.
13433 Cumulative number of frames detected as bottom field first using single-frame detection.
13435 @item multiple.current_frame
13436 Detected type of current frame using multiple-frame detection. One of:
13437 ``tff'' (top field first), ``bff'' (bottom field first),
13438 ``progressive'', or ``undetermined''
13441 Cumulative number of frames detected as bottom field first using multiple-frame detection.
13443 @item single.progressive
13444 Cumulative number of frames detected as progressive using single-frame detection.
13446 @item multiple.progressive
13447 Cumulative number of frames detected as progressive using multiple-frame detection.
13449 @item single.undetermined
13450 Cumulative number of frames that could not be classified using single-frame detection.
13452 @item multiple.undetermined
13453 Cumulative number of frames that could not be classified using multiple-frame detection.
13455 @item repeated.current_frame
13456 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
13458 @item repeated.neither
13459 Cumulative number of frames with no repeated field.
13462 Cumulative number of frames with the top field repeated from the previous frame's top field.
13464 @item repeated.bottom
13465 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
13468 The filter accepts the following options:
13472 Set interlacing threshold.
13474 Set progressive threshold.
13476 Threshold for repeated field detection.
13478 Number of frames after which a given frame's contribution to the
13479 statistics is halved (i.e., it contributes only 0.5 to its
13480 classification). The default of 0 means that all frames seen are given
13481 full weight of 1.0 forever.
13482 @item analyze_interlaced_flag
13483 When this is not 0 then idet will use the specified number of frames to determine
13484 if the interlaced flag is accurate, it will not count undetermined frames.
13485 If the flag is found to be accurate it will be used without any further
13486 computations, if it is found to be inaccurate it will be cleared without any
13487 further computations. This allows inserting the idet filter as a low computational
13488 method to clean up the interlaced flag
13493 Deinterleave or interleave fields.
13495 This filter allows one to process interlaced images fields without
13496 deinterlacing them. Deinterleaving splits the input frame into 2
13497 fields (so called half pictures). Odd lines are moved to the top
13498 half of the output image, even lines to the bottom half.
13499 You can process (filter) them independently and then re-interleave them.
13501 The filter accepts the following options:
13505 @item chroma_mode, c
13506 @item alpha_mode, a
13507 Available values for @var{luma_mode}, @var{chroma_mode} and
13508 @var{alpha_mode} are:
13514 @item deinterleave, d
13515 Deinterleave fields, placing one above the other.
13517 @item interleave, i
13518 Interleave fields. Reverse the effect of deinterleaving.
13520 Default value is @code{none}.
13522 @item luma_swap, ls
13523 @item chroma_swap, cs
13524 @item alpha_swap, as
13525 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
13528 @subsection Commands
13530 This filter supports the all above options as @ref{commands}.
13534 Apply inflate effect to the video.
13536 This filter replaces the pixel by the local(3x3) average by taking into account
13537 only values higher than the pixel.
13539 It accepts the following options:
13546 Limit the maximum change for each plane, default is 65535.
13547 If 0, plane will remain unchanged.
13550 @subsection Commands
13552 This filter supports the all above options as @ref{commands}.
13556 Simple interlacing filter from progressive contents. This interleaves upper (or
13557 lower) lines from odd frames with lower (or upper) lines from even frames,
13558 halving the frame rate and preserving image height.
13561 Original Original New Frame
13562 Frame 'j' Frame 'j+1' (tff)
13563 ========== =========== ==================
13564 Line 0 --------------------> Frame 'j' Line 0
13565 Line 1 Line 1 ----> Frame 'j+1' Line 1
13566 Line 2 ---------------------> Frame 'j' Line 2
13567 Line 3 Line 3 ----> Frame 'j+1' Line 3
13569 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
13572 It accepts the following optional parameters:
13576 This determines whether the interlaced frame is taken from the even
13577 (tff - default) or odd (bff) lines of the progressive frame.
13580 Vertical lowpass filter to avoid twitter interlacing and
13581 reduce moire patterns.
13585 Disable vertical lowpass filter
13588 Enable linear filter (default)
13591 Enable complex filter. This will slightly less reduce twitter and moire
13592 but better retain detail and subjective sharpness impression.
13599 Deinterlace input video by applying Donald Graft's adaptive kernel
13600 deinterling. Work on interlaced parts of a video to produce
13601 progressive frames.
13603 The description of the accepted parameters follows.
13607 Set the threshold which affects the filter's tolerance when
13608 determining if a pixel line must be processed. It must be an integer
13609 in the range [0,255] and defaults to 10. A value of 0 will result in
13610 applying the process on every pixels.
13613 Paint pixels exceeding the threshold value to white if set to 1.
13617 Set the fields order. Swap fields if set to 1, leave fields alone if
13621 Enable additional sharpening if set to 1. Default is 0.
13624 Enable twoway sharpening if set to 1. Default is 0.
13627 @subsection Examples
13631 Apply default values:
13633 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
13637 Enable additional sharpening:
13643 Paint processed pixels in white:
13650 Apply kirsch operator to input video stream.
13652 The filter accepts the following option:
13656 Set which planes will be processed, unprocessed planes will be copied.
13657 By default value 0xf, all planes will be processed.
13660 Set value which will be multiplied with filtered result.
13663 Set value which will be added to filtered result.
13666 @subsection Commands
13668 This filter supports the all above options as @ref{commands}.
13672 Slowly update darker pixels.
13674 This filter makes short flashes of light appear longer.
13675 This filter accepts the following options:
13679 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
13682 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
13685 @subsection Commands
13687 This filter supports the all above options as @ref{commands}.
13689 @section lenscorrection
13691 Correct radial lens distortion
13693 This filter can be used to correct for radial distortion as can result from the use
13694 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
13695 one can use tools available for example as part of opencv or simply trial-and-error.
13696 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
13697 and extract the k1 and k2 coefficients from the resulting matrix.
13699 Note that effectively the same filter is available in the open-source tools Krita and
13700 Digikam from the KDE project.
13702 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
13703 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
13704 brightness distribution, so you may want to use both filters together in certain
13705 cases, though you will have to take care of ordering, i.e. whether vignetting should
13706 be applied before or after lens correction.
13708 @subsection Options
13710 The filter accepts the following options:
13714 Relative x-coordinate of the focal point of the image, and thereby the center of the
13715 distortion. This value has a range [0,1] and is expressed as fractions of the image
13716 width. Default is 0.5.
13718 Relative y-coordinate of the focal point of the image, and thereby the center of the
13719 distortion. This value has a range [0,1] and is expressed as fractions of the image
13720 height. Default is 0.5.
13722 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
13723 no correction. Default is 0.
13725 Coefficient of the double quadratic correction term. This value has a range [-1,1].
13726 0 means no correction. Default is 0.
13728 Set interpolation type. Can be @code{nearest} or @code{bilinear}.
13729 Default is @code{nearest}.
13731 Specify the color of the unmapped pixels. For the syntax of this option,
13732 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13733 manual,ffmpeg-utils}. Default color is @code{black@@0}.
13736 The formula that generates the correction is:
13738 @var{r_src} = @var{r_tgt} * (1 + @var{k1} * (@var{r_tgt} / @var{r_0})^2 + @var{k2} * (@var{r_tgt} / @var{r_0})^4)
13740 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
13741 distances from the focal point in the source and target images, respectively.
13743 @subsection Commands
13745 This filter supports the all above options as @ref{commands}.
13749 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
13751 The @code{lensfun} filter requires the camera make, camera model, and lens model
13752 to apply the lens correction. The filter will load the lensfun database and
13753 query it to find the corresponding camera and lens entries in the database. As
13754 long as these entries can be found with the given options, the filter can
13755 perform corrections on frames. Note that incomplete strings will result in the
13756 filter choosing the best match with the given options, and the filter will
13757 output the chosen camera and lens models (logged with level "info"). You must
13758 provide the make, camera model, and lens model as they are required.
13760 The filter accepts the following options:
13764 The make of the camera (for example, "Canon"). This option is required.
13767 The model of the camera (for example, "Canon EOS 100D"). This option is
13771 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
13772 option is required.
13775 The type of correction to apply. The following values are valid options:
13779 Enables fixing lens vignetting.
13782 Enables fixing lens geometry. This is the default.
13785 Enables fixing chromatic aberrations.
13788 Enables fixing lens vignetting and lens geometry.
13791 Enables fixing lens vignetting and chromatic aberrations.
13794 Enables fixing both lens geometry and chromatic aberrations.
13797 Enables all possible corrections.
13801 The focal length of the image/video (zoom; expected constant for video). For
13802 example, a 18--55mm lens has focal length range of [18--55], so a value in that
13803 range should be chosen when using that lens. Default 18.
13806 The aperture of the image/video (expected constant for video). Note that
13807 aperture is only used for vignetting correction. Default 3.5.
13809 @item focus_distance
13810 The focus distance of the image/video (expected constant for video). Note that
13811 focus distance is only used for vignetting and only slightly affects the
13812 vignetting correction process. If unknown, leave it at the default value (which
13816 The scale factor which is applied after transformation. After correction the
13817 video is no longer necessarily rectangular. This parameter controls how much of
13818 the resulting image is visible. The value 0 means that a value will be chosen
13819 automatically such that there is little or no unmapped area in the output
13820 image. 1.0 means that no additional scaling is done. Lower values may result
13821 in more of the corrected image being visible, while higher values may avoid
13822 unmapped areas in the output.
13824 @item target_geometry
13825 The target geometry of the output image/video. The following values are valid
13829 @item rectilinear (default)
13832 @item equirectangular
13833 @item fisheye_orthographic
13834 @item fisheye_stereographic
13835 @item fisheye_equisolid
13836 @item fisheye_thoby
13839 Apply the reverse of image correction (instead of correcting distortion, apply
13842 @item interpolation
13843 The type of interpolation used when correcting distortion. The following values
13848 @item linear (default)
13853 @subsection Examples
13857 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
13858 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
13862 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8 -c:v h264 -b:v 8000k output.mov
13866 Apply the same as before, but only for the first 5 seconds of video.
13869 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8:enable='lte(t\,5)' -c:v h264 -b:v 8000k output.mov
13876 Obtain the VMAF (Video Multi-Method Assessment Fusion)
13877 score between two input videos.
13879 The obtained VMAF score is printed through the logging system.
13881 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
13882 After installing the library it can be enabled using:
13883 @code{./configure --enable-libvmaf}.
13884 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
13886 The filter has following options:
13890 Set the model path which is to be used for SVM.
13891 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
13894 Set the file path to be used to store logs.
13897 Set the format of the log file (csv, json or xml).
13899 @item enable_transform
13900 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
13901 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
13902 Default value: @code{false}
13905 Invokes the phone model which will generate VMAF scores higher than in the
13906 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
13907 Default value: @code{false}
13910 Enables computing psnr along with vmaf.
13911 Default value: @code{false}
13914 Enables computing ssim along with vmaf.
13915 Default value: @code{false}
13918 Enables computing ms_ssim along with vmaf.
13919 Default value: @code{false}
13922 Set the pool method to be used for computing vmaf.
13923 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
13926 Set number of threads to be used when computing vmaf.
13927 Default value: @code{0}, which makes use of all available logical processors.
13930 Set interval for frame subsampling used when computing vmaf.
13931 Default value: @code{1}
13933 @item enable_conf_interval
13934 Enables confidence interval.
13935 Default value: @code{false}
13938 This filter also supports the @ref{framesync} options.
13940 @subsection Examples
13943 On the below examples the input file @file{main.mpg} being processed is
13944 compared with the reference file @file{ref.mpg}.
13947 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
13951 Example with options:
13953 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
13957 Example with options and different containers:
13959 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]libvmaf=psnr=1:log_fmt=json" -f null -
13965 Limits the pixel components values to the specified range [min, max].
13967 The filter accepts the following options:
13971 Lower bound. Defaults to the lowest allowed value for the input.
13974 Upper bound. Defaults to the highest allowed value for the input.
13977 Specify which planes will be processed. Defaults to all available.
13980 @subsection Commands
13982 This filter supports the all above options as @ref{commands}.
13988 The filter accepts the following options:
13992 Set the number of loops. Setting this value to -1 will result in infinite loops.
13996 Set maximal size in number of frames. Default is 0.
13999 Set first frame of loop. Default is 0.
14002 @subsection Examples
14006 Loop single first frame infinitely:
14008 loop=loop=-1:size=1:start=0
14012 Loop single first frame 10 times:
14014 loop=loop=10:size=1:start=0
14018 Loop 10 first frames 5 times:
14020 loop=loop=5:size=10:start=0
14026 Apply a 1D LUT to an input video.
14028 The filter accepts the following options:
14032 Set the 1D LUT file name.
14034 Currently supported formats:
14043 Select interpolation mode.
14045 Available values are:
14049 Use values from the nearest defined point.
14051 Interpolate values using the linear interpolation.
14053 Interpolate values using the cosine interpolation.
14055 Interpolate values using the cubic interpolation.
14057 Interpolate values using the spline interpolation.
14061 @subsection Commands
14063 This filter supports the all above options as @ref{commands}.
14068 Apply a 3D LUT to an input video.
14070 The filter accepts the following options:
14074 Set the 3D LUT file name.
14076 Currently supported formats:
14090 Select interpolation mode.
14092 Available values are:
14096 Use values from the nearest defined point.
14098 Interpolate values using the 8 points defining a cube.
14100 Interpolate values using a tetrahedron.
14102 Interpolate values using a pyramid.
14104 Interpolate values using a prism.
14108 @subsection Commands
14110 This filter supports the @code{interp} option as @ref{commands}.
14114 Turn certain luma values into transparency.
14116 The filter accepts the following options:
14120 Set the luma which will be used as base for transparency.
14121 Default value is @code{0}.
14124 Set the range of luma values to be keyed out.
14125 Default value is @code{0.01}.
14128 Set the range of softness. Default value is @code{0}.
14129 Use this to control gradual transition from zero to full transparency.
14132 @subsection Commands
14133 This filter supports same @ref{commands} as options.
14134 The command accepts the same syntax of the corresponding option.
14136 If the specified expression is not valid, it is kept at its current
14139 @section lut, lutrgb, lutyuv
14141 Compute a look-up table for binding each pixel component input value
14142 to an output value, and apply it to the input video.
14144 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
14145 to an RGB input video.
14147 These filters accept the following parameters:
14150 set first pixel component expression
14152 set second pixel component expression
14154 set third pixel component expression
14156 set fourth pixel component expression, corresponds to the alpha component
14159 set red component expression
14161 set green component expression
14163 set blue component expression
14165 alpha component expression
14168 set Y/luminance component expression
14170 set U/Cb component expression
14172 set V/Cr component expression
14175 Each of them specifies the expression to use for computing the lookup table for
14176 the corresponding pixel component values.
14178 The exact component associated to each of the @var{c*} options depends on the
14181 The @var{lut} filter requires either YUV or RGB pixel formats in input,
14182 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
14184 The expressions can contain the following constants and functions:
14189 The input width and height.
14192 The input value for the pixel component.
14195 The input value, clipped to the @var{minval}-@var{maxval} range.
14198 The maximum value for the pixel component.
14201 The minimum value for the pixel component.
14204 The negated value for the pixel component value, clipped to the
14205 @var{minval}-@var{maxval} range; it corresponds to the expression
14206 "maxval-clipval+minval".
14209 The computed value in @var{val}, clipped to the
14210 @var{minval}-@var{maxval} range.
14212 @item gammaval(gamma)
14213 The computed gamma correction value of the pixel component value,
14214 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
14216 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
14220 All expressions default to "val".
14222 @subsection Commands
14224 This filter supports same @ref{commands} as options.
14226 @subsection Examples
14230 Negate input video:
14232 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
14233 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
14236 The above is the same as:
14238 lutrgb="r=negval:g=negval:b=negval"
14239 lutyuv="y=negval:u=negval:v=negval"
14249 Remove chroma components, turning the video into a graytone image:
14251 lutyuv="u=128:v=128"
14255 Apply a luma burning effect:
14261 Remove green and blue components:
14267 Set a constant alpha channel value on input:
14269 format=rgba,lutrgb=a="maxval-minval/2"
14273 Correct luminance gamma by a factor of 0.5:
14275 lutyuv=y=gammaval(0.5)
14279 Discard least significant bits of luma:
14281 lutyuv=y='bitand(val, 128+64+32)'
14285 Technicolor like effect:
14287 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
14291 @section lut2, tlut2
14293 The @code{lut2} filter takes two input streams and outputs one
14296 The @code{tlut2} (time lut2) filter takes two consecutive frames
14297 from one single stream.
14299 This filter accepts the following parameters:
14302 set first pixel component expression
14304 set second pixel component expression
14306 set third pixel component expression
14308 set fourth pixel component expression, corresponds to the alpha component
14311 set output bit depth, only available for @code{lut2} filter. By default is 0,
14312 which means bit depth is automatically picked from first input format.
14315 The @code{lut2} filter also supports the @ref{framesync} options.
14317 Each of them specifies the expression to use for computing the lookup table for
14318 the corresponding pixel component values.
14320 The exact component associated to each of the @var{c*} options depends on the
14323 The expressions can contain the following constants:
14328 The input width and height.
14331 The first input value for the pixel component.
14334 The second input value for the pixel component.
14337 The first input video bit depth.
14340 The second input video bit depth.
14343 All expressions default to "x".
14345 @subsection Commands
14347 This filter supports the all above options as @ref{commands} except option @code{d}.
14349 @subsection Examples
14353 Highlight differences between two RGB video streams:
14355 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1)'
14359 Highlight differences between two YUV video streams:
14361 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1)'
14365 Show max difference between two video streams:
14367 lut2='if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1)))'
14371 @section maskedclamp
14373 Clamp the first input stream with the second input and third input stream.
14375 Returns the value of first stream to be between second input
14376 stream - @code{undershoot} and third input stream + @code{overshoot}.
14378 This filter accepts the following options:
14381 Default value is @code{0}.
14384 Default value is @code{0}.
14387 Set which planes will be processed as bitmap, unprocessed planes will be
14388 copied from first stream.
14389 By default value 0xf, all planes will be processed.
14392 @subsection Commands
14394 This filter supports the all above options as @ref{commands}.
14398 Merge the second and third input stream into output stream using absolute differences
14399 between second input stream and first input stream and absolute difference between
14400 third input stream and first input stream. The picked value will be from second input
14401 stream if second absolute difference is greater than first one or from third input stream
14404 This filter accepts the following options:
14407 Set which planes will be processed as bitmap, unprocessed planes will be
14408 copied from first stream.
14409 By default value 0xf, all planes will be processed.
14412 @subsection Commands
14414 This filter supports the all above options as @ref{commands}.
14416 @section maskedmerge
14418 Merge the first input stream with the second input stream using per pixel
14419 weights in the third input stream.
14421 A value of 0 in the third stream pixel component means that pixel component
14422 from first stream is returned unchanged, while maximum value (eg. 255 for
14423 8-bit videos) means that pixel component from second stream is returned
14424 unchanged. Intermediate values define the amount of merging between both
14425 input stream's pixel components.
14427 This filter accepts the following options:
14430 Set which planes will be processed as bitmap, unprocessed planes will be
14431 copied from first stream.
14432 By default value 0xf, all planes will be processed.
14435 @subsection Commands
14437 This filter supports the all above options as @ref{commands}.
14441 Merge the second and third input stream into output stream using absolute differences
14442 between second input stream and first input stream and absolute difference between
14443 third input stream and first input stream. The picked value will be from second input
14444 stream if second absolute difference is less than first one or from third input stream
14447 This filter accepts the following options:
14450 Set which planes will be processed as bitmap, unprocessed planes will be
14451 copied from first stream.
14452 By default value 0xf, all planes will be processed.
14455 @subsection Commands
14457 This filter supports the all above options as @ref{commands}.
14459 @section maskedthreshold
14460 Pick pixels comparing absolute difference of two video streams with fixed
14463 If absolute difference between pixel component of first and second video
14464 stream is equal or lower than user supplied threshold than pixel component
14465 from first video stream is picked, otherwise pixel component from second
14466 video stream is picked.
14468 This filter accepts the following options:
14471 Set threshold used when picking pixels from absolute difference from two input
14475 Set which planes will be processed as bitmap, unprocessed planes will be
14476 copied from second stream.
14477 By default value 0xf, all planes will be processed.
14480 @subsection Commands
14482 This filter supports the all above options as @ref{commands}.
14485 Create mask from input video.
14487 For example it is useful to create motion masks after @code{tblend} filter.
14489 This filter accepts the following options:
14493 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
14496 Set high threshold. Any pixel component higher than this value will be set to max value
14497 allowed for current pixel format.
14500 Set planes to filter, by default all available planes are filtered.
14503 Fill all frame pixels with this value.
14506 Set max average pixel value for frame. If sum of all pixel components is higher that this
14507 average, output frame will be completely filled with value set by @var{fill} option.
14508 Typically useful for scene changes when used in combination with @code{tblend} filter.
14511 @subsection Commands
14513 This filter supports the all above options as @ref{commands}.
14517 Apply motion-compensation deinterlacing.
14519 It needs one field per frame as input and must thus be used together
14520 with yadif=1/3 or equivalent.
14522 This filter accepts the following options:
14525 Set the deinterlacing mode.
14527 It accepts one of the following values:
14532 use iterative motion estimation
14534 like @samp{slow}, but use multiple reference frames.
14536 Default value is @samp{fast}.
14539 Set the picture field parity assumed for the input video. It must be
14540 one of the following values:
14544 assume top field first
14546 assume bottom field first
14549 Default value is @samp{bff}.
14552 Set per-block quantization parameter (QP) used by the internal
14555 Higher values should result in a smoother motion vector field but less
14556 optimal individual vectors. Default value is 1.
14561 Pick median pixel from certain rectangle defined by radius.
14563 This filter accepts the following options:
14567 Set horizontal radius size. Default value is @code{1}.
14568 Allowed range is integer from 1 to 127.
14571 Set which planes to process. Default is @code{15}, which is all available planes.
14574 Set vertical radius size. Default value is @code{0}.
14575 Allowed range is integer from 0 to 127.
14576 If it is 0, value will be picked from horizontal @code{radius} option.
14579 Set median percentile. Default value is @code{0.5}.
14580 Default value of @code{0.5} will pick always median values, while @code{0} will pick
14581 minimum values, and @code{1} maximum values.
14584 @subsection Commands
14585 This filter supports same @ref{commands} as options.
14586 The command accepts the same syntax of the corresponding option.
14588 If the specified expression is not valid, it is kept at its current
14591 @section mergeplanes
14593 Merge color channel components from several video streams.
14595 The filter accepts up to 4 input streams, and merge selected input
14596 planes to the output video.
14598 This filter accepts the following options:
14601 Set input to output plane mapping. Default is @code{0}.
14603 The mappings is specified as a bitmap. It should be specified as a
14604 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
14605 mapping for the first plane of the output stream. 'A' sets the number of
14606 the input stream to use (from 0 to 3), and 'a' the plane number of the
14607 corresponding input to use (from 0 to 3). The rest of the mappings is
14608 similar, 'Bb' describes the mapping for the output stream second
14609 plane, 'Cc' describes the mapping for the output stream third plane and
14610 'Dd' describes the mapping for the output stream fourth plane.
14613 Set output pixel format. Default is @code{yuva444p}.
14616 @subsection Examples
14620 Merge three gray video streams of same width and height into single video stream:
14622 [a0][a1][a2]mergeplanes=0x001020:yuv444p
14626 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
14628 [a0][a1]mergeplanes=0x00010210:yuva444p
14632 Swap Y and A plane in yuva444p stream:
14634 format=yuva444p,mergeplanes=0x03010200:yuva444p
14638 Swap U and V plane in yuv420p stream:
14640 format=yuv420p,mergeplanes=0x000201:yuv420p
14644 Cast a rgb24 clip to yuv444p:
14646 format=rgb24,mergeplanes=0x000102:yuv444p
14652 Estimate and export motion vectors using block matching algorithms.
14653 Motion vectors are stored in frame side data to be used by other filters.
14655 This filter accepts the following options:
14658 Specify the motion estimation method. Accepts one of the following values:
14662 Exhaustive search algorithm.
14664 Three step search algorithm.
14666 Two dimensional logarithmic search algorithm.
14668 New three step search algorithm.
14670 Four step search algorithm.
14672 Diamond search algorithm.
14674 Hexagon-based search algorithm.
14676 Enhanced predictive zonal search algorithm.
14678 Uneven multi-hexagon search algorithm.
14680 Default value is @samp{esa}.
14683 Macroblock size. Default @code{16}.
14686 Search parameter. Default @code{7}.
14689 @section midequalizer
14691 Apply Midway Image Equalization effect using two video streams.
14693 Midway Image Equalization adjusts a pair of images to have the same
14694 histogram, while maintaining their dynamics as much as possible. It's
14695 useful for e.g. matching exposures from a pair of stereo cameras.
14697 This filter has two inputs and one output, which must be of same pixel format, but
14698 may be of different sizes. The output of filter is first input adjusted with
14699 midway histogram of both inputs.
14701 This filter accepts the following option:
14705 Set which planes to process. Default is @code{15}, which is all available planes.
14708 @section minterpolate
14710 Convert the video to specified frame rate using motion interpolation.
14712 This filter accepts the following options:
14715 Specify the output frame rate. This can be rational e.g. @code{60000/1001}. Frames are dropped if @var{fps} is lower than source fps. Default @code{60}.
14718 Motion interpolation mode. Following values are accepted:
14721 Duplicate previous or next frame for interpolating new ones.
14723 Blend source frames. Interpolated frame is mean of previous and next frames.
14725 Motion compensated interpolation. Following options are effective when this mode is selected:
14729 Motion compensation mode. Following values are accepted:
14732 Overlapped block motion compensation.
14734 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
14736 Default mode is @samp{obmc}.
14739 Motion estimation mode. Following values are accepted:
14742 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
14744 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
14746 Default mode is @samp{bilat}.
14749 The algorithm to be used for motion estimation. Following values are accepted:
14752 Exhaustive search algorithm.
14754 Three step search algorithm.
14756 Two dimensional logarithmic search algorithm.
14758 New three step search algorithm.
14760 Four step search algorithm.
14762 Diamond search algorithm.
14764 Hexagon-based search algorithm.
14766 Enhanced predictive zonal search algorithm.
14768 Uneven multi-hexagon search algorithm.
14770 Default algorithm is @samp{epzs}.
14773 Macroblock size. Default @code{16}.
14776 Motion estimation search parameter. Default @code{32}.
14779 Enable variable-size block motion compensation. Motion estimation is applied with smaller block sizes at object boundaries in order to make the them less blur. Default is @code{0} (disabled).
14784 Scene change detection method. Scene change leads motion vectors to be in random direction. Scene change detection replace interpolated frames by duplicate ones. May not be needed for other modes. Following values are accepted:
14787 Disable scene change detection.
14789 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
14791 Default method is @samp{fdiff}.
14793 @item scd_threshold
14794 Scene change detection threshold. Default is @code{10.}.
14799 Mix several video input streams into one video stream.
14801 A description of the accepted options follows.
14805 The number of inputs. If unspecified, it defaults to 2.
14808 Specify weight of each input video stream as sequence.
14809 Each weight is separated by space. If number of weights
14810 is smaller than number of @var{frames} last specified
14811 weight will be used for all remaining unset weights.
14814 Specify scale, if it is set it will be multiplied with sum
14815 of each weight multiplied with pixel values to give final destination
14816 pixel value. By default @var{scale} is auto scaled to sum of weights.
14819 Specify how end of stream is determined.
14822 The duration of the longest input. (default)
14825 The duration of the shortest input.
14828 The duration of the first input.
14832 @subsection Commands
14834 This filter supports the following commands:
14838 Syntax is same as option with same name.
14841 @section monochrome
14842 Convert video to gray using custom color filter.
14844 A description of the accepted options follows.
14848 Set the chroma blue spot. Allowed range is from -1 to 1.
14849 Default value is 0.
14852 Set the chroma red spot. Allowed range is from -1 to 1.
14853 Default value is 0.
14856 Set the color filter size. Allowed range is from .1 to 10.
14857 Default value is 1.
14860 Set the highlights strength. Allowed range is from 0 to 1.
14861 Default value is 0.
14864 @subsection Commands
14866 This filter supports the all above options as @ref{commands}.
14868 @section mpdecimate
14870 Drop frames that do not differ greatly from the previous frame in
14871 order to reduce frame rate.
14873 The main use of this filter is for very-low-bitrate encoding
14874 (e.g. streaming over dialup modem), but it could in theory be used for
14875 fixing movies that were inverse-telecined incorrectly.
14877 A description of the accepted options follows.
14881 Set the maximum number of consecutive frames which can be dropped (if
14882 positive), or the minimum interval between dropped frames (if
14883 negative). If the value is 0, the frame is dropped disregarding the
14884 number of previous sequentially dropped frames.
14886 Default value is 0.
14891 Set the dropping threshold values.
14893 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
14894 represent actual pixel value differences, so a threshold of 64
14895 corresponds to 1 unit of difference for each pixel, or the same spread
14896 out differently over the block.
14898 A frame is a candidate for dropping if no 8x8 blocks differ by more
14899 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
14900 meaning the whole image) differ by more than a threshold of @option{lo}.
14902 Default value for @option{hi} is 64*12, default value for @option{lo} is
14903 64*5, and default value for @option{frac} is 0.33.
14908 Obtain the MSAD (Mean Sum of Absolute Differences) between two input videos.
14910 This filter takes two input videos.
14912 Both input videos must have the same resolution and pixel format for
14913 this filter to work correctly. Also it assumes that both inputs
14914 have the same number of frames, which are compared one by one.
14916 The obtained per component, average, min and max MSAD is printed through
14917 the logging system.
14919 The filter stores the calculated MSAD of each frame in frame metadata.
14921 In the below example the input file @file{main.mpg} being processed is compared
14922 with the reference file @file{ref.mpg}.
14925 ffmpeg -i main.mpg -i ref.mpg -lavfi msad -f null -
14930 Negate (invert) the input video.
14932 It accepts the following option:
14937 With value 1, it negates the alpha component, if present. Default value is 0.
14940 @subsection Commands
14942 This filter supports same @ref{commands} as options.
14947 Denoise frames using Non-Local Means algorithm.
14949 Each pixel is adjusted by looking for other pixels with similar contexts. This
14950 context similarity is defined by comparing their surrounding patches of size
14951 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
14954 Note that the research area defines centers for patches, which means some
14955 patches will be made of pixels outside that research area.
14957 The filter accepts the following options.
14961 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
14964 Set patch size. Default is 7. Must be odd number in range [0, 99].
14967 Same as @option{p} but for chroma planes.
14969 The default value is @var{0} and means automatic.
14972 Set research size. Default is 15. Must be odd number in range [0, 99].
14975 Same as @option{r} but for chroma planes.
14977 The default value is @var{0} and means automatic.
14982 Deinterlace video using neural network edge directed interpolation.
14984 This filter accepts the following options:
14988 Mandatory option, without binary file filter can not work.
14989 Currently file can be found here:
14990 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
14993 Set which frames to deinterlace, by default it is @code{all}.
14994 Can be @code{all} or @code{interlaced}.
14997 Set mode of operation.
14999 Can be one of the following:
15003 Use frame flags, both fields.
15005 Use frame flags, single field.
15007 Use top field only.
15009 Use bottom field only.
15011 Use both fields, top first.
15013 Use both fields, bottom first.
15017 Set which planes to process, by default filter process all frames.
15020 Set size of local neighborhood around each pixel, used by the predictor neural
15023 Can be one of the following:
15036 Set the number of neurons in predictor neural network.
15037 Can be one of the following:
15048 Controls the number of different neural network predictions that are blended
15049 together to compute the final output value. Can be @code{fast}, default or
15053 Set which set of weights to use in the predictor.
15054 Can be one of the following:
15058 weights trained to minimize absolute error
15060 weights trained to minimize squared error
15064 Controls whether or not the prescreener neural network is used to decide
15065 which pixels should be processed by the predictor neural network and which
15066 can be handled by simple cubic interpolation.
15067 The prescreener is trained to know whether cubic interpolation will be
15068 sufficient for a pixel or whether it should be predicted by the predictor nn.
15069 The computational complexity of the prescreener nn is much less than that of
15070 the predictor nn. Since most pixels can be handled by cubic interpolation,
15071 using the prescreener generally results in much faster processing.
15072 The prescreener is pretty accurate, so the difference between using it and not
15073 using it is almost always unnoticeable.
15075 Can be one of the following:
15085 Default is @code{new}.
15088 @subsection Commands
15089 This filter supports same @ref{commands} as options, excluding @var{weights} option.
15093 Force libavfilter not to use any of the specified pixel formats for the
15094 input to the next filter.
15096 It accepts the following parameters:
15100 A '|'-separated list of pixel format names, such as
15101 pix_fmts=yuv420p|monow|rgb24".
15105 @subsection Examples
15109 Force libavfilter to use a format different from @var{yuv420p} for the
15110 input to the vflip filter:
15112 noformat=pix_fmts=yuv420p,vflip
15116 Convert the input video to any of the formats not contained in the list:
15118 noformat=yuv420p|yuv444p|yuv410p
15124 Add noise on video input frame.
15126 The filter accepts the following options:
15134 Set noise seed for specific pixel component or all pixel components in case
15135 of @var{all_seed}. Default value is @code{123457}.
15137 @item all_strength, alls
15138 @item c0_strength, c0s
15139 @item c1_strength, c1s
15140 @item c2_strength, c2s
15141 @item c3_strength, c3s
15142 Set noise strength for specific pixel component or all pixel components in case
15143 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
15145 @item all_flags, allf
15146 @item c0_flags, c0f
15147 @item c1_flags, c1f
15148 @item c2_flags, c2f
15149 @item c3_flags, c3f
15150 Set pixel component flags or set flags for all components if @var{all_flags}.
15151 Available values for component flags are:
15154 averaged temporal noise (smoother)
15156 mix random noise with a (semi)regular pattern
15158 temporal noise (noise pattern changes between frames)
15160 uniform noise (gaussian otherwise)
15164 @subsection Examples
15166 Add temporal and uniform noise to input video:
15168 noise=alls=20:allf=t+u
15173 Normalize RGB video (aka histogram stretching, contrast stretching).
15174 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
15176 For each channel of each frame, the filter computes the input range and maps
15177 it linearly to the user-specified output range. The output range defaults
15178 to the full dynamic range from pure black to pure white.
15180 Temporal smoothing can be used on the input range to reduce flickering (rapid
15181 changes in brightness) caused when small dark or bright objects enter or leave
15182 the scene. This is similar to the auto-exposure (automatic gain control) on a
15183 video camera, and, like a video camera, it may cause a period of over- or
15184 under-exposure of the video.
15186 The R,G,B channels can be normalized independently, which may cause some
15187 color shifting, or linked together as a single channel, which prevents
15188 color shifting. Linked normalization preserves hue. Independent normalization
15189 does not, so it can be used to remove some color casts. Independent and linked
15190 normalization can be combined in any ratio.
15192 The normalize filter accepts the following options:
15197 Colors which define the output range. The minimum input value is mapped to
15198 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
15199 The defaults are black and white respectively. Specifying white for
15200 @var{blackpt} and black for @var{whitept} will give color-inverted,
15201 normalized video. Shades of grey can be used to reduce the dynamic range
15202 (contrast). Specifying saturated colors here can create some interesting
15206 The number of previous frames to use for temporal smoothing. The input range
15207 of each channel is smoothed using a rolling average over the current frame
15208 and the @var{smoothing} previous frames. The default is 0 (no temporal
15212 Controls the ratio of independent (color shifting) channel normalization to
15213 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
15214 independent. Defaults to 1.0 (fully independent).
15217 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
15218 expensive no-op. Defaults to 1.0 (full strength).
15222 @subsection Commands
15223 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
15224 The command accepts the same syntax of the corresponding option.
15226 If the specified expression is not valid, it is kept at its current
15229 @subsection Examples
15231 Stretch video contrast to use the full dynamic range, with no temporal
15232 smoothing; may flicker depending on the source content:
15234 normalize=blackpt=black:whitept=white:smoothing=0
15237 As above, but with 50 frames of temporal smoothing; flicker should be
15238 reduced, depending on the source content:
15240 normalize=blackpt=black:whitept=white:smoothing=50
15243 As above, but with hue-preserving linked channel normalization:
15245 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
15248 As above, but with half strength:
15250 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
15253 Map the darkest input color to red, the brightest input color to cyan:
15255 normalize=blackpt=red:whitept=cyan
15260 Pass the video source unchanged to the output.
15263 Optical Character Recognition
15265 This filter uses Tesseract for optical character recognition. To enable
15266 compilation of this filter, you need to configure FFmpeg with
15267 @code{--enable-libtesseract}.
15269 It accepts the following options:
15273 Set datapath to tesseract data. Default is to use whatever was
15274 set at installation.
15277 Set language, default is "eng".
15280 Set character whitelist.
15283 Set character blacklist.
15286 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
15287 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
15291 Apply a video transform using libopencv.
15293 To enable this filter, install the libopencv library and headers and
15294 configure FFmpeg with @code{--enable-libopencv}.
15296 It accepts the following parameters:
15301 The name of the libopencv filter to apply.
15303 @item filter_params
15304 The parameters to pass to the libopencv filter. If not specified, the default
15305 values are assumed.
15309 Refer to the official libopencv documentation for more precise
15311 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
15313 Several libopencv filters are supported; see the following subsections.
15318 Dilate an image by using a specific structuring element.
15319 It corresponds to the libopencv function @code{cvDilate}.
15321 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
15323 @var{struct_el} represents a structuring element, and has the syntax:
15324 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
15326 @var{cols} and @var{rows} represent the number of columns and rows of
15327 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
15328 point, and @var{shape} the shape for the structuring element. @var{shape}
15329 must be "rect", "cross", "ellipse", or "custom".
15331 If the value for @var{shape} is "custom", it must be followed by a
15332 string of the form "=@var{filename}". The file with name
15333 @var{filename} is assumed to represent a binary image, with each
15334 printable character corresponding to a bright pixel. When a custom
15335 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
15336 or columns and rows of the read file are assumed instead.
15338 The default value for @var{struct_el} is "3x3+0x0/rect".
15340 @var{nb_iterations} specifies the number of times the transform is
15341 applied to the image, and defaults to 1.
15345 # Use the default values
15348 # Dilate using a structuring element with a 5x5 cross, iterating two times
15349 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
15351 # Read the shape from the file diamond.shape, iterating two times.
15352 # The file diamond.shape may contain a pattern of characters like this
15358 # The specified columns and rows are ignored
15359 # but the anchor point coordinates are not
15360 ocv=dilate:0x0+2x2/custom=diamond.shape|2
15365 Erode an image by using a specific structuring element.
15366 It corresponds to the libopencv function @code{cvErode}.
15368 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
15369 with the same syntax and semantics as the @ref{dilate} filter.
15373 Smooth the input video.
15375 The filter takes the following parameters:
15376 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
15378 @var{type} is the type of smooth filter to apply, and must be one of
15379 the following values: "blur", "blur_no_scale", "median", "gaussian",
15380 or "bilateral". The default value is "gaussian".
15382 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
15383 depends on the smooth type. @var{param1} and
15384 @var{param2} accept integer positive values or 0. @var{param3} and
15385 @var{param4} accept floating point values.
15387 The default value for @var{param1} is 3. The default value for the
15388 other parameters is 0.
15390 These parameters correspond to the parameters assigned to the
15391 libopencv function @code{cvSmooth}.
15393 @section oscilloscope
15395 2D Video Oscilloscope.
15397 Useful to measure spatial impulse, step responses, chroma delays, etc.
15399 It accepts the following parameters:
15403 Set scope center x position.
15406 Set scope center y position.
15409 Set scope size, relative to frame diagonal.
15412 Set scope tilt/rotation.
15418 Set trace center x position.
15421 Set trace center y position.
15424 Set trace width, relative to width of frame.
15427 Set trace height, relative to height of frame.
15430 Set which components to trace. By default it traces first three components.
15433 Draw trace grid. By default is enabled.
15436 Draw some statistics. By default is enabled.
15439 Draw scope. By default is enabled.
15442 @subsection Commands
15443 This filter supports same @ref{commands} as options.
15444 The command accepts the same syntax of the corresponding option.
15446 If the specified expression is not valid, it is kept at its current
15449 @subsection Examples
15453 Inspect full first row of video frame.
15455 oscilloscope=x=0.5:y=0:s=1
15459 Inspect full last row of video frame.
15461 oscilloscope=x=0.5:y=1:s=1
15465 Inspect full 5th line of video frame of height 1080.
15467 oscilloscope=x=0.5:y=5/1080:s=1
15471 Inspect full last column of video frame.
15473 oscilloscope=x=1:y=0.5:s=1:t=1
15481 Overlay one video on top of another.
15483 It takes two inputs and has one output. The first input is the "main"
15484 video on which the second input is overlaid.
15486 It accepts the following parameters:
15488 A description of the accepted options follows.
15493 Set the expression for the x and y coordinates of the overlaid video
15494 on the main video. Default value is "0" for both expressions. In case
15495 the expression is invalid, it is set to a huge value (meaning that the
15496 overlay will not be displayed within the output visible area).
15499 See @ref{framesync}.
15502 Set when the expressions for @option{x}, and @option{y} are evaluated.
15504 It accepts the following values:
15507 only evaluate expressions once during the filter initialization or
15508 when a command is processed
15511 evaluate expressions for each incoming frame
15514 Default value is @samp{frame}.
15517 See @ref{framesync}.
15520 Set the format for the output video.
15522 It accepts the following values:
15525 force YUV420 output
15528 force YUV420p10 output
15531 force YUV422 output
15534 force YUV422p10 output
15537 force YUV444 output
15540 force packed RGB output
15543 force planar RGB output
15546 automatically pick format
15549 Default value is @samp{yuv420}.
15552 See @ref{framesync}.
15555 Set format of alpha of the overlaid video, it can be @var{straight} or
15556 @var{premultiplied}. Default is @var{straight}.
15559 The @option{x}, and @option{y} expressions can contain the following
15565 The main input width and height.
15569 The overlay input width and height.
15573 The computed values for @var{x} and @var{y}. They are evaluated for
15578 horizontal and vertical chroma subsample values of the output
15579 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
15583 the number of input frame, starting from 0
15586 the position in the file of the input frame, NAN if unknown
15589 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
15593 This filter also supports the @ref{framesync} options.
15595 Note that the @var{n}, @var{pos}, @var{t} variables are available only
15596 when evaluation is done @emph{per frame}, and will evaluate to NAN
15597 when @option{eval} is set to @samp{init}.
15599 Be aware that frames are taken from each input video in timestamp
15600 order, hence, if their initial timestamps differ, it is a good idea
15601 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
15602 have them begin in the same zero timestamp, as the example for
15603 the @var{movie} filter does.
15605 You can chain together more overlays but you should test the
15606 efficiency of such approach.
15608 @subsection Commands
15610 This filter supports the following commands:
15614 Modify the x and y of the overlay input.
15615 The command accepts the same syntax of the corresponding option.
15617 If the specified expression is not valid, it is kept at its current
15621 @subsection Examples
15625 Draw the overlay at 10 pixels from the bottom right corner of the main
15628 overlay=main_w-overlay_w-10:main_h-overlay_h-10
15631 Using named options the example above becomes:
15633 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
15637 Insert a transparent PNG logo in the bottom left corner of the input,
15638 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
15640 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
15644 Insert 2 different transparent PNG logos (second logo on bottom
15645 right corner) using the @command{ffmpeg} tool:
15647 ffmpeg -i input -i logo1 -i logo2 -filter_complex 'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
15651 Add a transparent color layer on top of the main video; @code{WxH}
15652 must specify the size of the main input to the overlay filter:
15654 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
15658 Play an original video and a filtered version (here with the deshake
15659 filter) side by side using the @command{ffplay} tool:
15661 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
15664 The above command is the same as:
15666 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
15670 Make a sliding overlay appearing from the left to the right top part of the
15671 screen starting since time 2:
15673 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
15677 Compose output by putting two input videos side to side:
15679 ffmpeg -i left.avi -i right.avi -filter_complex "
15680 nullsrc=size=200x100 [background];
15681 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
15682 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
15683 [background][left] overlay=shortest=1 [background+left];
15684 [background+left][right] overlay=shortest=1:x=100 [left+right]
15689 Mask 10-20 seconds of a video by applying the delogo filter to a section
15691 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
15692 -vf '[in]split[split_main][split_delogo];[split_delogo]trim=start=360:end=371,delogo=0:0:640:480[delogoed];[split_main][delogoed]overlay=eof_action=pass[out]'
15697 Chain several overlays in cascade:
15699 nullsrc=s=200x200 [bg];
15700 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
15701 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
15702 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
15703 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
15704 [in3] null, [mid2] overlay=100:100 [out0]
15709 @anchor{overlay_cuda}
15710 @section overlay_cuda
15712 Overlay one video on top of another.
15714 This is the CUDA variant of the @ref{overlay} filter.
15715 It only accepts CUDA frames. The underlying input pixel formats have to match.
15717 It takes two inputs and has one output. The first input is the "main"
15718 video on which the second input is overlaid.
15720 It accepts the following parameters:
15725 Set the x and y coordinates of the overlaid video on the main video.
15726 Default value is "0" for both expressions.
15729 See @ref{framesync}.
15732 See @ref{framesync}.
15735 See @ref{framesync}.
15739 This filter also supports the @ref{framesync} options.
15743 Apply Overcomplete Wavelet denoiser.
15745 The filter accepts the following options:
15751 Larger depth values will denoise lower frequency components more, but
15752 slow down filtering.
15754 Must be an int in the range 8-16, default is @code{8}.
15756 @item luma_strength, ls
15759 Must be a double value in the range 0-1000, default is @code{1.0}.
15761 @item chroma_strength, cs
15762 Set chroma strength.
15764 Must be a double value in the range 0-1000, default is @code{1.0}.
15770 Add paddings to the input image, and place the original input at the
15771 provided @var{x}, @var{y} coordinates.
15773 It accepts the following parameters:
15778 Specify an expression for the size of the output image with the
15779 paddings added. If the value for @var{width} or @var{height} is 0, the
15780 corresponding input size is used for the output.
15782 The @var{width} expression can reference the value set by the
15783 @var{height} expression, and vice versa.
15785 The default value of @var{width} and @var{height} is 0.
15789 Specify the offsets to place the input image at within the padded area,
15790 with respect to the top/left border of the output image.
15792 The @var{x} expression can reference the value set by the @var{y}
15793 expression, and vice versa.
15795 The default value of @var{x} and @var{y} is 0.
15797 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
15798 so the input image is centered on the padded area.
15801 Specify the color of the padded area. For the syntax of this option,
15802 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15803 manual,ffmpeg-utils}.
15805 The default value of @var{color} is "black".
15808 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
15810 It accepts the following values:
15814 Only evaluate expressions once during the filter initialization or when
15815 a command is processed.
15818 Evaluate expressions for each incoming frame.
15822 Default value is @samp{init}.
15825 Pad to aspect instead to a resolution.
15829 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
15830 options are expressions containing the following constants:
15835 The input video width and height.
15839 These are the same as @var{in_w} and @var{in_h}.
15843 The output width and height (the size of the padded area), as
15844 specified by the @var{width} and @var{height} expressions.
15848 These are the same as @var{out_w} and @var{out_h}.
15852 The x and y offsets as specified by the @var{x} and @var{y}
15853 expressions, or NAN if not yet specified.
15856 same as @var{iw} / @var{ih}
15859 input sample aspect ratio
15862 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
15866 The horizontal and vertical chroma subsample values. For example for the
15867 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15870 @subsection Examples
15874 Add paddings with the color "violet" to the input video. The output video
15875 size is 640x480, and the top-left corner of the input video is placed at
15878 pad=640:480:0:40:violet
15881 The example above is equivalent to the following command:
15883 pad=width=640:height=480:x=0:y=40:color=violet
15887 Pad the input to get an output with dimensions increased by 3/2,
15888 and put the input video at the center of the padded area:
15890 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
15894 Pad the input to get a squared output with size equal to the maximum
15895 value between the input width and height, and put the input video at
15896 the center of the padded area:
15898 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
15902 Pad the input to get a final w/h ratio of 16:9:
15904 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
15908 In case of anamorphic video, in order to set the output display aspect
15909 correctly, it is necessary to use @var{sar} in the expression,
15910 according to the relation:
15912 (ih * X / ih) * sar = output_dar
15913 X = output_dar / sar
15916 Thus the previous example needs to be modified to:
15918 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
15922 Double the output size and put the input video in the bottom-right
15923 corner of the output padded area:
15925 pad="2*iw:2*ih:ow-iw:oh-ih"
15929 @anchor{palettegen}
15930 @section palettegen
15932 Generate one palette for a whole video stream.
15934 It accepts the following options:
15938 Set the maximum number of colors to quantize in the palette.
15939 Note: the palette will still contain 256 colors; the unused palette entries
15942 @item reserve_transparent
15943 Create a palette of 255 colors maximum and reserve the last one for
15944 transparency. Reserving the transparency color is useful for GIF optimization.
15945 If not set, the maximum of colors in the palette will be 256. You probably want
15946 to disable this option for a standalone image.
15949 @item transparency_color
15950 Set the color that will be used as background for transparency.
15953 Set statistics mode.
15955 It accepts the following values:
15958 Compute full frame histograms.
15960 Compute histograms only for the part that differs from previous frame. This
15961 might be relevant to give more importance to the moving part of your input if
15962 the background is static.
15964 Compute new histogram for each frame.
15967 Default value is @var{full}.
15970 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
15971 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
15972 color quantization of the palette. This information is also visible at
15973 @var{info} logging level.
15975 @subsection Examples
15979 Generate a representative palette of a given video using @command{ffmpeg}:
15981 ffmpeg -i input.mkv -vf palettegen palette.png
15985 @section paletteuse
15987 Use a palette to downsample an input video stream.
15989 The filter takes two inputs: one video stream and a palette. The palette must
15990 be a 256 pixels image.
15992 It accepts the following options:
15996 Select dithering mode. Available algorithms are:
15999 Ordered 8x8 bayer dithering (deterministic)
16001 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
16002 Note: this dithering is sometimes considered "wrong" and is included as a
16004 @item floyd_steinberg
16005 Floyd and Steingberg dithering (error diffusion)
16007 Frankie Sierra dithering v2 (error diffusion)
16009 Frankie Sierra dithering v2 "Lite" (error diffusion)
16012 Default is @var{sierra2_4a}.
16015 When @var{bayer} dithering is selected, this option defines the scale of the
16016 pattern (how much the crosshatch pattern is visible). A low value means more
16017 visible pattern for less banding, and higher value means less visible pattern
16018 at the cost of more banding.
16020 The option must be an integer value in the range [0,5]. Default is @var{2}.
16023 If set, define the zone to process
16027 Only the changing rectangle will be reprocessed. This is similar to GIF
16028 cropping/offsetting compression mechanism. This option can be useful for speed
16029 if only a part of the image is changing, and has use cases such as limiting the
16030 scope of the error diffusal @option{dither} to the rectangle that bounds the
16031 moving scene (it leads to more deterministic output if the scene doesn't change
16032 much, and as a result less moving noise and better GIF compression).
16035 Default is @var{none}.
16038 Take new palette for each output frame.
16040 @item alpha_threshold
16041 Sets the alpha threshold for transparency. Alpha values above this threshold
16042 will be treated as completely opaque, and values below this threshold will be
16043 treated as completely transparent.
16045 The option must be an integer value in the range [0,255]. Default is @var{128}.
16048 @subsection Examples
16052 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
16053 using @command{ffmpeg}:
16055 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
16059 @section perspective
16061 Correct perspective of video not recorded perpendicular to the screen.
16063 A description of the accepted parameters follows.
16074 Set coordinates expression for top left, top right, bottom left and bottom right corners.
16075 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
16076 If the @code{sense} option is set to @code{source}, then the specified points will be sent
16077 to the corners of the destination. If the @code{sense} option is set to @code{destination},
16078 then the corners of the source will be sent to the specified coordinates.
16080 The expressions can use the following variables:
16085 the width and height of video frame.
16089 Output frame count.
16092 @item interpolation
16093 Set interpolation for perspective correction.
16095 It accepts the following values:
16101 Default value is @samp{linear}.
16104 Set interpretation of coordinate options.
16106 It accepts the following values:
16110 Send point in the source specified by the given coordinates to
16111 the corners of the destination.
16113 @item 1, destination
16115 Send the corners of the source to the point in the destination specified
16116 by the given coordinates.
16118 Default value is @samp{source}.
16122 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
16124 It accepts the following values:
16127 only evaluate expressions once during the filter initialization or
16128 when a command is processed
16131 evaluate expressions for each incoming frame
16134 Default value is @samp{init}.
16139 Delay interlaced video by one field time so that the field order changes.
16141 The intended use is to fix PAL movies that have been captured with the
16142 opposite field order to the film-to-video transfer.
16144 A description of the accepted parameters follows.
16150 It accepts the following values:
16153 Capture field order top-first, transfer bottom-first.
16154 Filter will delay the bottom field.
16157 Capture field order bottom-first, transfer top-first.
16158 Filter will delay the top field.
16161 Capture and transfer with the same field order. This mode only exists
16162 for the documentation of the other options to refer to, but if you
16163 actually select it, the filter will faithfully do nothing.
16166 Capture field order determined automatically by field flags, transfer
16168 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
16169 basis using field flags. If no field information is available,
16170 then this works just like @samp{u}.
16173 Capture unknown or varying, transfer opposite.
16174 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
16175 analyzing the images and selecting the alternative that produces best
16176 match between the fields.
16179 Capture top-first, transfer unknown or varying.
16180 Filter selects among @samp{t} and @samp{p} using image analysis.
16183 Capture bottom-first, transfer unknown or varying.
16184 Filter selects among @samp{b} and @samp{p} using image analysis.
16187 Capture determined by field flags, transfer unknown or varying.
16188 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
16189 image analysis. If no field information is available, then this works just
16190 like @samp{U}. This is the default mode.
16193 Both capture and transfer unknown or varying.
16194 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
16198 @subsection Commands
16200 This filter supports the all above options as @ref{commands}.
16202 @section photosensitivity
16203 Reduce various flashes in video, so to help users with epilepsy.
16205 It accepts the following options:
16208 Set how many frames to use when filtering. Default is 30.
16211 Set detection threshold factor. Default is 1.
16215 Set how many pixels to skip when sampling frames. Default is 1.
16216 Allowed range is from 1 to 1024.
16219 Leave frames unchanged. Default is disabled.
16222 @section pixdesctest
16224 Pixel format descriptor test filter, mainly useful for internal
16225 testing. The output video should be equal to the input video.
16229 format=monow, pixdesctest
16232 can be used to test the monowhite pixel format descriptor definition.
16236 Display sample values of color channels. Mainly useful for checking color
16237 and levels. Minimum supported resolution is 640x480.
16239 The filters accept the following options:
16243 Set scope X position, relative offset on X axis.
16246 Set scope Y position, relative offset on Y axis.
16255 Set window opacity. This window also holds statistics about pixel area.
16258 Set window X position, relative offset on X axis.
16261 Set window Y position, relative offset on Y axis.
16264 @subsection Commands
16266 This filter supports same @ref{commands} as options.
16270 Enable the specified chain of postprocessing subfilters using libpostproc. This
16271 library should be automatically selected with a GPL build (@code{--enable-gpl}).
16272 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
16273 Each subfilter and some options have a short and a long name that can be used
16274 interchangeably, i.e. dr/dering are the same.
16276 The filters accept the following options:
16280 Set postprocessing subfilters string.
16283 All subfilters share common options to determine their scope:
16287 Honor the quality commands for this subfilter.
16290 Do chrominance filtering, too (default).
16293 Do luminance filtering only (no chrominance).
16296 Do chrominance filtering only (no luminance).
16299 These options can be appended after the subfilter name, separated by a '|'.
16301 Available subfilters are:
16304 @item hb/hdeblock[|difference[|flatness]]
16305 Horizontal deblocking filter
16308 Difference factor where higher values mean more deblocking (default: @code{32}).
16310 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16313 @item vb/vdeblock[|difference[|flatness]]
16314 Vertical deblocking filter
16317 Difference factor where higher values mean more deblocking (default: @code{32}).
16319 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16322 @item ha/hadeblock[|difference[|flatness]]
16323 Accurate horizontal deblocking filter
16326 Difference factor where higher values mean more deblocking (default: @code{32}).
16328 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16331 @item va/vadeblock[|difference[|flatness]]
16332 Accurate vertical deblocking filter
16335 Difference factor where higher values mean more deblocking (default: @code{32}).
16337 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16341 The horizontal and vertical deblocking filters share the difference and
16342 flatness values so you cannot set different horizontal and vertical
16346 @item h1/x1hdeblock
16347 Experimental horizontal deblocking filter
16349 @item v1/x1vdeblock
16350 Experimental vertical deblocking filter
16355 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
16358 larger -> stronger filtering
16360 larger -> stronger filtering
16362 larger -> stronger filtering
16365 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
16368 Stretch luminance to @code{0-255}.
16371 @item lb/linblenddeint
16372 Linear blend deinterlacing filter that deinterlaces the given block by
16373 filtering all lines with a @code{(1 2 1)} filter.
16375 @item li/linipoldeint
16376 Linear interpolating deinterlacing filter that deinterlaces the given block by
16377 linearly interpolating every second line.
16379 @item ci/cubicipoldeint
16380 Cubic interpolating deinterlacing filter deinterlaces the given block by
16381 cubically interpolating every second line.
16383 @item md/mediandeint
16384 Median deinterlacing filter that deinterlaces the given block by applying a
16385 median filter to every second line.
16387 @item fd/ffmpegdeint
16388 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
16389 second line with a @code{(-1 4 2 4 -1)} filter.
16392 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
16393 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
16395 @item fq/forceQuant[|quantizer]
16396 Overrides the quantizer table from the input with the constant quantizer you
16404 Default pp filter combination (@code{hb|a,vb|a,dr|a})
16407 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
16410 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
16413 @subsection Examples
16417 Apply horizontal and vertical deblocking, deringing and automatic
16418 brightness/contrast:
16424 Apply default filters without brightness/contrast correction:
16430 Apply default filters and temporal denoiser:
16432 pp=default/tmpnoise|1|2|3
16436 Apply deblocking on luminance only, and switch vertical deblocking on or off
16437 automatically depending on available CPU time:
16444 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
16445 similar to spp = 6 with 7 point DCT, where only the center sample is
16448 The filter accepts the following options:
16452 Force a constant quantization parameter. It accepts an integer in range
16453 0 to 63. If not set, the filter will use the QP from the video stream
16457 Set thresholding mode. Available modes are:
16461 Set hard thresholding.
16463 Set soft thresholding (better de-ringing effect, but likely blurrier).
16465 Set medium thresholding (good results, default).
16469 @section premultiply
16470 Apply alpha premultiply effect to input video stream using first plane
16471 of second stream as alpha.
16473 Both streams must have same dimensions and same pixel format.
16475 The filter accepts the following option:
16479 Set which planes will be processed, unprocessed planes will be copied.
16480 By default value 0xf, all planes will be processed.
16483 Do not require 2nd input for processing, instead use alpha plane from input stream.
16487 Apply prewitt operator to input video stream.
16489 The filter accepts the following option:
16493 Set which planes will be processed, unprocessed planes will be copied.
16494 By default value 0xf, all planes will be processed.
16497 Set value which will be multiplied with filtered result.
16500 Set value which will be added to filtered result.
16503 @subsection Commands
16505 This filter supports the all above options as @ref{commands}.
16507 @section pseudocolor
16509 Alter frame colors in video with pseudocolors.
16511 This filter accepts the following options:
16515 set pixel first component expression
16518 set pixel second component expression
16521 set pixel third component expression
16524 set pixel fourth component expression, corresponds to the alpha component
16527 set component to use as base for altering colors
16530 Pick one of built-in LUTs. By default is set to none.
16547 Set opacity of output colors. Allowed range is from 0 to 1.
16548 Default value is set to 1.
16551 Each of the expression options specifies the expression to use for computing
16552 the lookup table for the corresponding pixel component values.
16554 The expressions can contain the following constants and functions:
16559 The input width and height.
16562 The input value for the pixel component.
16564 @item ymin, umin, vmin, amin
16565 The minimum allowed component value.
16567 @item ymax, umax, vmax, amax
16568 The maximum allowed component value.
16571 All expressions default to "val".
16573 @subsection Commands
16575 This filter supports the all above options as @ref{commands}.
16577 @subsection Examples
16581 Change too high luma values to gradient:
16583 pseudocolor="'if(between(val,ymax,amax),lerp(ymin,ymax,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(umax,umin,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(vmin,vmax,(val-ymax)/(amax-ymax)),-1):-1'"
16589 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
16590 Ratio) between two input videos.
16592 This filter takes in input two input videos, the first input is
16593 considered the "main" source and is passed unchanged to the
16594 output. The second input is used as a "reference" video for computing
16597 Both video inputs must have the same resolution and pixel format for
16598 this filter to work correctly. Also it assumes that both inputs
16599 have the same number of frames, which are compared one by one.
16601 The obtained average PSNR is printed through the logging system.
16603 The filter stores the accumulated MSE (mean squared error) of each
16604 frame, and at the end of the processing it is averaged across all frames
16605 equally, and the following formula is applied to obtain the PSNR:
16608 PSNR = 10*log10(MAX^2/MSE)
16611 Where MAX is the average of the maximum values of each component of the
16614 The description of the accepted parameters follows.
16617 @item stats_file, f
16618 If specified the filter will use the named file to save the PSNR of
16619 each individual frame. When filename equals "-" the data is sent to
16622 @item stats_version
16623 Specifies which version of the stats file format to use. Details of
16624 each format are written below.
16625 Default value is 1.
16627 @item stats_add_max
16628 Determines whether the max value is output to the stats log.
16629 Default value is 0.
16630 Requires stats_version >= 2. If this is set and stats_version < 2,
16631 the filter will return an error.
16634 This filter also supports the @ref{framesync} options.
16636 The file printed if @var{stats_file} is selected, contains a sequence of
16637 key/value pairs of the form @var{key}:@var{value} for each compared
16640 If a @var{stats_version} greater than 1 is specified, a header line precedes
16641 the list of per-frame-pair stats, with key value pairs following the frame
16642 format with the following parameters:
16645 @item psnr_log_version
16646 The version of the log file format. Will match @var{stats_version}.
16649 A comma separated list of the per-frame-pair parameters included in
16653 A description of each shown per-frame-pair parameter follows:
16657 sequential number of the input frame, starting from 1
16660 Mean Square Error pixel-by-pixel average difference of the compared
16661 frames, averaged over all the image components.
16663 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
16664 Mean Square Error pixel-by-pixel average difference of the compared
16665 frames for the component specified by the suffix.
16667 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
16668 Peak Signal to Noise ratio of the compared frames for the component
16669 specified by the suffix.
16671 @item max_avg, max_y, max_u, max_v
16672 Maximum allowed value for each channel, and average over all
16676 @subsection Examples
16681 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16682 [main][ref] psnr="stats_file=stats.log" [out]
16685 On this example the input file being processed is compared with the
16686 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
16687 is stored in @file{stats.log}.
16690 Another example with different containers:
16692 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]psnr" -f null -
16699 Pulldown reversal (inverse telecine) filter, capable of handling mixed
16700 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
16703 The pullup filter is designed to take advantage of future context in making
16704 its decisions. This filter is stateless in the sense that it does not lock
16705 onto a pattern to follow, but it instead looks forward to the following
16706 fields in order to identify matches and rebuild progressive frames.
16708 To produce content with an even framerate, insert the fps filter after
16709 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
16710 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
16712 The filter accepts the following options:
16719 These options set the amount of "junk" to ignore at the left, right, top, and
16720 bottom of the image, respectively. Left and right are in units of 8 pixels,
16721 while top and bottom are in units of 2 lines.
16722 The default is 8 pixels on each side.
16725 Set the strict breaks. Setting this option to 1 will reduce the chances of
16726 filter generating an occasional mismatched frame, but it may also cause an
16727 excessive number of frames to be dropped during high motion sequences.
16728 Conversely, setting it to -1 will make filter match fields more easily.
16729 This may help processing of video where there is slight blurring between
16730 the fields, but may also cause there to be interlaced frames in the output.
16731 Default value is @code{0}.
16734 Set the metric plane to use. It accepts the following values:
16740 Use chroma blue plane.
16743 Use chroma red plane.
16746 This option may be set to use chroma plane instead of the default luma plane
16747 for doing filter's computations. This may improve accuracy on very clean
16748 source material, but more likely will decrease accuracy, especially if there
16749 is chroma noise (rainbow effect) or any grayscale video.
16750 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
16751 load and make pullup usable in realtime on slow machines.
16754 For best results (without duplicated frames in the output file) it is
16755 necessary to change the output frame rate. For example, to inverse
16756 telecine NTSC input:
16758 ffmpeg -i input -vf pullup -r 24000/1001 ...
16763 Change video quantization parameters (QP).
16765 The filter accepts the following option:
16769 Set expression for quantization parameter.
16772 The expression is evaluated through the eval API and can contain, among others,
16773 the following constants:
16777 1 if index is not 129, 0 otherwise.
16780 Sequential index starting from -129 to 128.
16783 @subsection Examples
16787 Some equation like:
16795 Flush video frames from internal cache of frames into a random order.
16796 No frame is discarded.
16797 Inspired by @ref{frei0r} nervous filter.
16801 Set size in number of frames of internal cache, in range from @code{2} to
16802 @code{512}. Default is @code{30}.
16805 Set seed for random number generator, must be an integer included between
16806 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16807 less than @code{0}, the filter will try to use a good random seed on a
16811 @section readeia608
16813 Read closed captioning (EIA-608) information from the top lines of a video frame.
16815 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
16816 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
16817 with EIA-608 data (starting from 0). A description of each metadata value follows:
16820 @item lavfi.readeia608.X.cc
16821 The two bytes stored as EIA-608 data (printed in hexadecimal).
16823 @item lavfi.readeia608.X.line
16824 The number of the line on which the EIA-608 data was identified and read.
16827 This filter accepts the following options:
16831 Set the line to start scanning for EIA-608 data. Default is @code{0}.
16834 Set the line to end scanning for EIA-608 data. Default is @code{29}.
16837 Set the ratio of width reserved for sync code detection.
16838 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
16841 Enable checking the parity bit. In the event of a parity error, the filter will output
16842 @code{0x00} for that character. Default is false.
16845 Lowpass lines prior to further processing. Default is enabled.
16848 @subsection Commands
16850 This filter supports the all above options as @ref{commands}.
16852 @subsection Examples
16856 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
16858 ffprobe -f lavfi -i movie=captioned_video.mov,readeia608 -show_entries frame=pkt_pts_time:frame_tags=lavfi.readeia608.0.cc,lavfi.readeia608.1.cc -of csv
16864 Read vertical interval timecode (VITC) information from the top lines of a
16867 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
16868 timecode value, if a valid timecode has been detected. Further metadata key
16869 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
16870 timecode data has been found or not.
16872 This filter accepts the following options:
16876 Set the maximum number of lines to scan for VITC data. If the value is set to
16877 @code{-1} the full video frame is scanned. Default is @code{45}.
16880 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
16881 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
16884 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
16885 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
16888 @subsection Examples
16892 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
16893 draw @code{--:--:--:--} as a placeholder:
16895 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
16901 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
16903 Destination pixel at position (X, Y) will be picked from source (x, y) position
16904 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
16905 value for pixel will be used for destination pixel.
16907 Xmap and Ymap input video streams must be of same dimensions. Output video stream
16908 will have Xmap/Ymap video stream dimensions.
16909 Xmap and Ymap input video streams are 16bit depth, single channel.
16913 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
16914 Default is @code{color}.
16917 Specify the color of the unmapped pixels. For the syntax of this option,
16918 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16919 manual,ffmpeg-utils}. Default color is @code{black}.
16922 @section removegrain
16924 The removegrain filter is a spatial denoiser for progressive video.
16928 Set mode for the first plane.
16931 Set mode for the second plane.
16934 Set mode for the third plane.
16937 Set mode for the fourth plane.
16940 Range of mode is from 0 to 24. Description of each mode follows:
16944 Leave input plane unchanged. Default.
16947 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
16950 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
16953 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
16956 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
16957 This is equivalent to a median filter.
16960 Line-sensitive clipping giving the minimal change.
16963 Line-sensitive clipping, intermediate.
16966 Line-sensitive clipping, intermediate.
16969 Line-sensitive clipping, intermediate.
16972 Line-sensitive clipping on a line where the neighbours pixels are the closest.
16975 Replaces the target pixel with the closest neighbour.
16978 [1 2 1] horizontal and vertical kernel blur.
16984 Bob mode, interpolates top field from the line where the neighbours
16985 pixels are the closest.
16988 Bob mode, interpolates bottom field from the line where the neighbours
16989 pixels are the closest.
16992 Bob mode, interpolates top field. Same as 13 but with a more complicated
16993 interpolation formula.
16996 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
16997 interpolation formula.
17000 Clips the pixel with the minimum and maximum of respectively the maximum and
17001 minimum of each pair of opposite neighbour pixels.
17004 Line-sensitive clipping using opposite neighbours whose greatest distance from
17005 the current pixel is minimal.
17008 Replaces the pixel with the average of its 8 neighbours.
17011 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
17014 Clips pixels using the averages of opposite neighbour.
17017 Same as mode 21 but simpler and faster.
17020 Small edge and halo removal, but reputed useless.
17026 @section removelogo
17028 Suppress a TV station logo, using an image file to determine which
17029 pixels comprise the logo. It works by filling in the pixels that
17030 comprise the logo with neighboring pixels.
17032 The filter accepts the following options:
17036 Set the filter bitmap file, which can be any image format supported by
17037 libavformat. The width and height of the image file must match those of the
17038 video stream being processed.
17041 Pixels in the provided bitmap image with a value of zero are not
17042 considered part of the logo, non-zero pixels are considered part of
17043 the logo. If you use white (255) for the logo and black (0) for the
17044 rest, you will be safe. For making the filter bitmap, it is
17045 recommended to take a screen capture of a black frame with the logo
17046 visible, and then using a threshold filter followed by the erode
17047 filter once or twice.
17049 If needed, little splotches can be fixed manually. Remember that if
17050 logo pixels are not covered, the filter quality will be much
17051 reduced. Marking too many pixels as part of the logo does not hurt as
17052 much, but it will increase the amount of blurring needed to cover over
17053 the image and will destroy more information than necessary, and extra
17054 pixels will slow things down on a large logo.
17056 @section repeatfields
17058 This filter uses the repeat_field flag from the Video ES headers and hard repeats
17059 fields based on its value.
17063 Reverse a video clip.
17065 Warning: This filter requires memory to buffer the entire clip, so trimming
17068 @subsection Examples
17072 Take the first 5 seconds of a clip, and reverse it.
17079 Shift R/G/B/A pixels horizontally and/or vertically.
17081 The filter accepts the following options:
17084 Set amount to shift red horizontally.
17086 Set amount to shift red vertically.
17088 Set amount to shift green horizontally.
17090 Set amount to shift green vertically.
17092 Set amount to shift blue horizontally.
17094 Set amount to shift blue vertically.
17096 Set amount to shift alpha horizontally.
17098 Set amount to shift alpha vertically.
17100 Set edge mode, can be @var{smear}, default, or @var{warp}.
17103 @subsection Commands
17105 This filter supports the all above options as @ref{commands}.
17108 Apply roberts cross operator to input video stream.
17110 The filter accepts the following option:
17114 Set which planes will be processed, unprocessed planes will be copied.
17115 By default value 0xf, all planes will be processed.
17118 Set value which will be multiplied with filtered result.
17121 Set value which will be added to filtered result.
17124 @subsection Commands
17126 This filter supports the all above options as @ref{commands}.
17130 Rotate video by an arbitrary angle expressed in radians.
17132 The filter accepts the following options:
17134 A description of the optional parameters follows.
17137 Set an expression for the angle by which to rotate the input video
17138 clockwise, expressed as a number of radians. A negative value will
17139 result in a counter-clockwise rotation. By default it is set to "0".
17141 This expression is evaluated for each frame.
17144 Set the output width expression, default value is "iw".
17145 This expression is evaluated just once during configuration.
17148 Set the output height expression, default value is "ih".
17149 This expression is evaluated just once during configuration.
17152 Enable bilinear interpolation if set to 1, a value of 0 disables
17153 it. Default value is 1.
17156 Set the color used to fill the output area not covered by the rotated
17157 image. For the general syntax of this option, check the
17158 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17159 If the special value "none" is selected then no
17160 background is printed (useful for example if the background is never shown).
17162 Default value is "black".
17165 The expressions for the angle and the output size can contain the
17166 following constants and functions:
17170 sequential number of the input frame, starting from 0. It is always NAN
17171 before the first frame is filtered.
17174 time in seconds of the input frame, it is set to 0 when the filter is
17175 configured. It is always NAN before the first frame is filtered.
17179 horizontal and vertical chroma subsample values. For example for the
17180 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17184 the input video width and height
17188 the output width and height, that is the size of the padded area as
17189 specified by the @var{width} and @var{height} expressions
17193 the minimal width/height required for completely containing the input
17194 video rotated by @var{a} radians.
17196 These are only available when computing the @option{out_w} and
17197 @option{out_h} expressions.
17200 @subsection Examples
17204 Rotate the input by PI/6 radians clockwise:
17210 Rotate the input by PI/6 radians counter-clockwise:
17216 Rotate the input by 45 degrees clockwise:
17222 Apply a constant rotation with period T, starting from an angle of PI/3:
17224 rotate=PI/3+2*PI*t/T
17228 Make the input video rotation oscillating with a period of T
17229 seconds and an amplitude of A radians:
17231 rotate=A*sin(2*PI/T*t)
17235 Rotate the video, output size is chosen so that the whole rotating
17236 input video is always completely contained in the output:
17238 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
17242 Rotate the video, reduce the output size so that no background is ever
17245 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
17249 @subsection Commands
17251 The filter supports the following commands:
17255 Set the angle expression.
17256 The command accepts the same syntax of the corresponding option.
17258 If the specified expression is not valid, it is kept at its current
17264 Apply Shape Adaptive Blur.
17266 The filter accepts the following options:
17269 @item luma_radius, lr
17270 Set luma blur filter strength, must be a value in range 0.1-4.0, default
17271 value is 1.0. A greater value will result in a more blurred image, and
17272 in slower processing.
17274 @item luma_pre_filter_radius, lpfr
17275 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
17278 @item luma_strength, ls
17279 Set luma maximum difference between pixels to still be considered, must
17280 be a value in the 0.1-100.0 range, default value is 1.0.
17282 @item chroma_radius, cr
17283 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
17284 greater value will result in a more blurred image, and in slower
17287 @item chroma_pre_filter_radius, cpfr
17288 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
17290 @item chroma_strength, cs
17291 Set chroma maximum difference between pixels to still be considered,
17292 must be a value in the -0.9-100.0 range.
17295 Each chroma option value, if not explicitly specified, is set to the
17296 corresponding luma option value.
17301 Scale (resize) the input video, using the libswscale library.
17303 The scale filter forces the output display aspect ratio to be the same
17304 of the input, by changing the output sample aspect ratio.
17306 If the input image format is different from the format requested by
17307 the next filter, the scale filter will convert the input to the
17310 @subsection Options
17311 The filter accepts the following options, or any of the options
17312 supported by the libswscale scaler.
17314 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
17315 the complete list of scaler options.
17320 Set the output video dimension expression. Default value is the input
17323 If the @var{width} or @var{w} value is 0, the input width is used for
17324 the output. If the @var{height} or @var{h} value is 0, the input height
17325 is used for the output.
17327 If one and only one of the values is -n with n >= 1, the scale filter
17328 will use a value that maintains the aspect ratio of the input image,
17329 calculated from the other specified dimension. After that it will,
17330 however, make sure that the calculated dimension is divisible by n and
17331 adjust the value if necessary.
17333 If both values are -n with n >= 1, the behavior will be identical to
17334 both values being set to 0 as previously detailed.
17336 See below for the list of accepted constants for use in the dimension
17340 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
17344 Only evaluate expressions once during the filter initialization or when a command is processed.
17347 Evaluate expressions for each incoming frame.
17351 Default value is @samp{init}.
17355 Set the interlacing mode. It accepts the following values:
17359 Force interlaced aware scaling.
17362 Do not apply interlaced scaling.
17365 Select interlaced aware scaling depending on whether the source frames
17366 are flagged as interlaced or not.
17369 Default value is @samp{0}.
17372 Set libswscale scaling flags. See
17373 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17374 complete list of values. If not explicitly specified the filter applies
17378 @item param0, param1
17379 Set libswscale input parameters for scaling algorithms that need them. See
17380 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17381 complete documentation. If not explicitly specified the filter applies
17387 Set the video size. For the syntax of this option, check the
17388 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17390 @item in_color_matrix
17391 @item out_color_matrix
17392 Set in/output YCbCr color space type.
17394 This allows the autodetected value to be overridden as well as allows forcing
17395 a specific value used for the output and encoder.
17397 If not specified, the color space type depends on the pixel format.
17403 Choose automatically.
17406 Format conforming to International Telecommunication Union (ITU)
17407 Recommendation BT.709.
17410 Set color space conforming to the United States Federal Communications
17411 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
17416 Set color space conforming to:
17420 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
17423 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
17426 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
17431 Set color space conforming to SMPTE ST 240:1999.
17434 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
17439 Set in/output YCbCr sample range.
17441 This allows the autodetected value to be overridden as well as allows forcing
17442 a specific value used for the output and encoder. If not specified, the
17443 range depends on the pixel format. Possible values:
17447 Choose automatically.
17450 Set full range (0-255 in case of 8-bit luma).
17452 @item mpeg/limited/tv
17453 Set "MPEG" range (16-235 in case of 8-bit luma).
17456 @item force_original_aspect_ratio
17457 Enable decreasing or increasing output video width or height if necessary to
17458 keep the original aspect ratio. Possible values:
17462 Scale the video as specified and disable this feature.
17465 The output video dimensions will automatically be decreased if needed.
17468 The output video dimensions will automatically be increased if needed.
17472 One useful instance of this option is that when you know a specific device's
17473 maximum allowed resolution, you can use this to limit the output video to
17474 that, while retaining the aspect ratio. For example, device A allows
17475 1280x720 playback, and your video is 1920x800. Using this option (set it to
17476 decrease) and specifying 1280x720 to the command line makes the output
17479 Please note that this is a different thing than specifying -1 for @option{w}
17480 or @option{h}, you still need to specify the output resolution for this option
17483 @item force_divisible_by
17484 Ensures that both the output dimensions, width and height, are divisible by the
17485 given integer when used together with @option{force_original_aspect_ratio}. This
17486 works similar to using @code{-n} in the @option{w} and @option{h} options.
17488 This option respects the value set for @option{force_original_aspect_ratio},
17489 increasing or decreasing the resolution accordingly. The video's aspect ratio
17490 may be slightly modified.
17492 This option can be handy if you need to have a video fit within or exceed
17493 a defined resolution using @option{force_original_aspect_ratio} but also have
17494 encoder restrictions on width or height divisibility.
17498 The values of the @option{w} and @option{h} options are expressions
17499 containing the following constants:
17504 The input width and height
17508 These are the same as @var{in_w} and @var{in_h}.
17512 The output (scaled) width and height
17516 These are the same as @var{out_w} and @var{out_h}
17519 The same as @var{iw} / @var{ih}
17522 input sample aspect ratio
17525 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17529 horizontal and vertical input chroma subsample values. For example for the
17530 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17534 horizontal and vertical output chroma subsample values. For example for the
17535 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17538 The (sequential) number of the input frame, starting from 0.
17539 Only available with @code{eval=frame}.
17542 The presentation timestamp of the input frame, expressed as a number of
17543 seconds. Only available with @code{eval=frame}.
17546 The position (byte offset) of the frame in the input stream, or NaN if
17547 this information is unavailable and/or meaningless (for example in case of synthetic video).
17548 Only available with @code{eval=frame}.
17551 @subsection Examples
17555 Scale the input video to a size of 200x100
17560 This is equivalent to:
17571 Specify a size abbreviation for the output size:
17576 which can also be written as:
17582 Scale the input to 2x:
17584 scale=w=2*iw:h=2*ih
17588 The above is the same as:
17590 scale=2*in_w:2*in_h
17594 Scale the input to 2x with forced interlaced scaling:
17596 scale=2*iw:2*ih:interl=1
17600 Scale the input to half size:
17602 scale=w=iw/2:h=ih/2
17606 Increase the width, and set the height to the same size:
17612 Seek Greek harmony:
17619 Increase the height, and set the width to 3/2 of the height:
17621 scale=w=3/2*oh:h=3/5*ih
17625 Increase the size, making the size a multiple of the chroma
17628 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
17632 Increase the width to a maximum of 500 pixels,
17633 keeping the same aspect ratio as the input:
17635 scale=w='min(500\, iw*3/2):h=-1'
17639 Make pixels square by combining scale and setsar:
17641 scale='trunc(ih*dar):ih',setsar=1/1
17645 Make pixels square by combining scale and setsar,
17646 making sure the resulting resolution is even (required by some codecs):
17648 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
17652 @subsection Commands
17654 This filter supports the following commands:
17658 Set the output video dimension expression.
17659 The command accepts the same syntax of the corresponding option.
17661 If the specified expression is not valid, it is kept at its current
17667 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
17668 format conversion on CUDA video frames. Setting the output width and height
17669 works in the same way as for the @var{scale} filter.
17671 The following additional options are accepted:
17674 The pixel format of the output CUDA frames. If set to the string "same" (the
17675 default), the input format will be kept. Note that automatic format negotiation
17676 and conversion is not yet supported for hardware frames
17679 The interpolation algorithm used for resizing. One of the following:
17686 @item cubic2p_bspline
17687 2-parameter cubic (B=1, C=0)
17689 @item cubic2p_catmullrom
17690 2-parameter cubic (B=0, C=1/2)
17692 @item cubic2p_b05c03
17693 2-parameter cubic (B=1/2, C=3/10)
17701 @item force_original_aspect_ratio
17702 Enable decreasing or increasing output video width or height if necessary to
17703 keep the original aspect ratio. Possible values:
17707 Scale the video as specified and disable this feature.
17710 The output video dimensions will automatically be decreased if needed.
17713 The output video dimensions will automatically be increased if needed.
17717 One useful instance of this option is that when you know a specific device's
17718 maximum allowed resolution, you can use this to limit the output video to
17719 that, while retaining the aspect ratio. For example, device A allows
17720 1280x720 playback, and your video is 1920x800. Using this option (set it to
17721 decrease) and specifying 1280x720 to the command line makes the output
17724 Please note that this is a different thing than specifying -1 for @option{w}
17725 or @option{h}, you still need to specify the output resolution for this option
17728 @item force_divisible_by
17729 Ensures that both the output dimensions, width and height, are divisible by the
17730 given integer when used together with @option{force_original_aspect_ratio}. This
17731 works similar to using @code{-n} in the @option{w} and @option{h} options.
17733 This option respects the value set for @option{force_original_aspect_ratio},
17734 increasing or decreasing the resolution accordingly. The video's aspect ratio
17735 may be slightly modified.
17737 This option can be handy if you need to have a video fit within or exceed
17738 a defined resolution using @option{force_original_aspect_ratio} but also have
17739 encoder restrictions on width or height divisibility.
17745 Scale (resize) the input video, based on a reference video.
17747 See the scale filter for available options, scale2ref supports the same but
17748 uses the reference video instead of the main input as basis. scale2ref also
17749 supports the following additional constants for the @option{w} and
17750 @option{h} options:
17755 The main input video's width and height
17758 The same as @var{main_w} / @var{main_h}
17761 The main input video's sample aspect ratio
17763 @item main_dar, mdar
17764 The main input video's display aspect ratio. Calculated from
17765 @code{(main_w / main_h) * main_sar}.
17769 The main input video's horizontal and vertical chroma subsample values.
17770 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
17774 The (sequential) number of the main input frame, starting from 0.
17775 Only available with @code{eval=frame}.
17778 The presentation timestamp of the main input frame, expressed as a number of
17779 seconds. Only available with @code{eval=frame}.
17782 The position (byte offset) of the frame in the main input stream, or NaN if
17783 this information is unavailable and/or meaningless (for example in case of synthetic video).
17784 Only available with @code{eval=frame}.
17787 @subsection Examples
17791 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
17793 'scale2ref[b][a];[a][b]overlay'
17797 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
17799 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
17803 @subsection Commands
17805 This filter supports the following commands:
17809 Set the output video dimension expression.
17810 The command accepts the same syntax of the corresponding option.
17812 If the specified expression is not valid, it is kept at its current
17817 Scroll input video horizontally and/or vertically by constant speed.
17819 The filter accepts the following options:
17821 @item horizontal, h
17822 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
17823 Negative values changes scrolling direction.
17826 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
17827 Negative values changes scrolling direction.
17830 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
17833 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
17836 @subsection Commands
17838 This filter supports the following @ref{commands}:
17840 @item horizontal, h
17841 Set the horizontal scrolling speed.
17843 Set the vertical scrolling speed.
17849 Detect video scene change.
17851 This filter sets frame metadata with mafd between frame, the scene score, and
17852 forward the frame to the next filter, so they can use these metadata to detect
17853 scene change or others.
17855 In addition, this filter logs a message and sets frame metadata when it detects
17856 a scene change by @option{threshold}.
17858 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
17860 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
17861 to detect scene change.
17863 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
17864 detect scene change with @option{threshold}.
17866 The filter accepts the following options:
17870 Set the scene change detection threshold as a percentage of maximum change. Good
17871 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
17874 Default value is @code{10.}.
17877 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
17878 You can enable it if you want to get snapshot of scene change frames only.
17881 @anchor{selectivecolor}
17882 @section selectivecolor
17884 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
17885 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
17886 by the "purity" of the color (that is, how saturated it already is).
17888 This filter is similar to the Adobe Photoshop Selective Color tool.
17890 The filter accepts the following options:
17893 @item correction_method
17894 Select color correction method.
17896 Available values are:
17899 Specified adjustments are applied "as-is" (added/subtracted to original pixel
17902 Specified adjustments are relative to the original component value.
17904 Default is @code{absolute}.
17906 Adjustments for red pixels (pixels where the red component is the maximum)
17908 Adjustments for yellow pixels (pixels where the blue component is the minimum)
17910 Adjustments for green pixels (pixels where the green component is the maximum)
17912 Adjustments for cyan pixels (pixels where the red component is the minimum)
17914 Adjustments for blue pixels (pixels where the blue component is the maximum)
17916 Adjustments for magenta pixels (pixels where the green component is the minimum)
17918 Adjustments for white pixels (pixels where all components are greater than 128)
17920 Adjustments for all pixels except pure black and pure white
17922 Adjustments for black pixels (pixels where all components are lesser than 128)
17924 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
17927 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
17928 4 space separated floating point adjustment values in the [-1,1] range,
17929 respectively to adjust the amount of cyan, magenta, yellow and black for the
17930 pixels of its range.
17932 @subsection Examples
17936 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
17937 increase magenta by 27% in blue areas:
17939 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
17943 Use a Photoshop selective color preset:
17945 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
17949 @anchor{separatefields}
17950 @section separatefields
17952 The @code{separatefields} takes a frame-based video input and splits
17953 each frame into its components fields, producing a new half height clip
17954 with twice the frame rate and twice the frame count.
17956 This filter use field-dominance information in frame to decide which
17957 of each pair of fields to place first in the output.
17958 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
17960 @section setdar, setsar
17962 The @code{setdar} filter sets the Display Aspect Ratio for the filter
17965 This is done by changing the specified Sample (aka Pixel) Aspect
17966 Ratio, according to the following equation:
17968 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
17971 Keep in mind that the @code{setdar} filter does not modify the pixel
17972 dimensions of the video frame. Also, the display aspect ratio set by
17973 this filter may be changed by later filters in the filterchain,
17974 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
17977 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
17978 the filter output video.
17980 Note that as a consequence of the application of this filter, the
17981 output display aspect ratio will change according to the equation
17984 Keep in mind that the sample aspect ratio set by the @code{setsar}
17985 filter may be changed by later filters in the filterchain, e.g. if
17986 another "setsar" or a "setdar" filter is applied.
17988 It accepts the following parameters:
17991 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
17992 Set the aspect ratio used by the filter.
17994 The parameter can be a floating point number string, an expression, or
17995 a string of the form @var{num}:@var{den}, where @var{num} and
17996 @var{den} are the numerator and denominator of the aspect ratio. If
17997 the parameter is not specified, it is assumed the value "0".
17998 In case the form "@var{num}:@var{den}" is used, the @code{:} character
18002 Set the maximum integer value to use for expressing numerator and
18003 denominator when reducing the expressed aspect ratio to a rational.
18004 Default value is @code{100}.
18008 The parameter @var{sar} is an expression containing
18009 the following constants:
18013 These are approximated values for the mathematical constants e
18014 (Euler's number), pi (Greek pi), and phi (the golden ratio).
18017 The input width and height.
18020 These are the same as @var{w} / @var{h}.
18023 The input sample aspect ratio.
18026 The input display aspect ratio. It is the same as
18027 (@var{w} / @var{h}) * @var{sar}.
18030 Horizontal and vertical chroma subsample values. For example, for the
18031 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18034 @subsection Examples
18039 To change the display aspect ratio to 16:9, specify one of the following:
18046 To change the sample aspect ratio to 10:11, specify:
18052 To set a display aspect ratio of 16:9, and specify a maximum integer value of
18053 1000 in the aspect ratio reduction, use the command:
18055 setdar=ratio=16/9:max=1000
18063 Force field for the output video frame.
18065 The @code{setfield} filter marks the interlace type field for the
18066 output frames. It does not change the input frame, but only sets the
18067 corresponding property, which affects how the frame is treated by
18068 following filters (e.g. @code{fieldorder} or @code{yadif}).
18070 The filter accepts the following options:
18075 Available values are:
18079 Keep the same field property.
18082 Mark the frame as bottom-field-first.
18085 Mark the frame as top-field-first.
18088 Mark the frame as progressive.
18095 Force frame parameter for the output video frame.
18097 The @code{setparams} filter marks interlace and color range for the
18098 output frames. It does not change the input frame, but only sets the
18099 corresponding property, which affects how the frame is treated by
18104 Available values are:
18108 Keep the same field property (default).
18111 Mark the frame as bottom-field-first.
18114 Mark the frame as top-field-first.
18117 Mark the frame as progressive.
18121 Available values are:
18125 Keep the same color range property (default).
18127 @item unspecified, unknown
18128 Mark the frame as unspecified color range.
18130 @item limited, tv, mpeg
18131 Mark the frame as limited range.
18133 @item full, pc, jpeg
18134 Mark the frame as full range.
18137 @item color_primaries
18138 Set the color primaries.
18139 Available values are:
18143 Keep the same color primaries property (default).
18160 Set the color transfer.
18161 Available values are:
18165 Keep the same color trc property (default).
18187 Set the colorspace.
18188 Available values are:
18192 Keep the same colorspace property (default).
18205 @item chroma-derived-nc
18206 @item chroma-derived-c
18212 Apply shear transform to input video.
18214 This filter supports the following options:
18218 Shear factor in X-direction. Default value is 0.
18219 Allowed range is from -2 to 2.
18222 Shear factor in Y-direction. Default value is 0.
18223 Allowed range is from -2 to 2.
18226 Set the color used to fill the output area not covered by the transformed
18227 video. For the general syntax of this option, check the
18228 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18229 If the special value "none" is selected then no
18230 background is printed (useful for example if the background is never shown).
18232 Default value is "black".
18235 Set interpolation type. Can be @code{bilinear} or @code{nearest}. Default is @code{bilinear}.
18238 @subsection Commands
18240 This filter supports the all above options as @ref{commands}.
18244 Show a line containing various information for each input video frame.
18245 The input video is not modified.
18247 This filter supports the following options:
18251 Calculate checksums of each plane. By default enabled.
18254 The shown line contains a sequence of key/value pairs of the form
18255 @var{key}:@var{value}.
18257 The following values are shown in the output:
18261 The (sequential) number of the input frame, starting from 0.
18264 The Presentation TimeStamp of the input frame, expressed as a number of
18265 time base units. The time base unit depends on the filter input pad.
18268 The Presentation TimeStamp of the input frame, expressed as a number of
18272 The position of the frame in the input stream, or -1 if this information is
18273 unavailable and/or meaningless (for example in case of synthetic video).
18276 The pixel format name.
18279 The sample aspect ratio of the input frame, expressed in the form
18280 @var{num}/@var{den}.
18283 The size of the input frame. For the syntax of this option, check the
18284 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18287 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
18288 for bottom field first).
18291 This is 1 if the frame is a key frame, 0 otherwise.
18294 The picture type of the input frame ("I" for an I-frame, "P" for a
18295 P-frame, "B" for a B-frame, or "?" for an unknown type).
18296 Also refer to the documentation of the @code{AVPictureType} enum and of
18297 the @code{av_get_picture_type_char} function defined in
18298 @file{libavutil/avutil.h}.
18301 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
18303 @item plane_checksum
18304 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
18305 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
18308 The mean value of pixels in each plane of the input frame, expressed in the form
18309 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
18312 The standard deviation of pixel values in each plane of the input frame, expressed
18313 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
18317 @section showpalette
18319 Displays the 256 colors palette of each frame. This filter is only relevant for
18320 @var{pal8} pixel format frames.
18322 It accepts the following option:
18326 Set the size of the box used to represent one palette color entry. Default is
18327 @code{30} (for a @code{30x30} pixel box).
18330 @section shuffleframes
18332 Reorder and/or duplicate and/or drop video frames.
18334 It accepts the following parameters:
18338 Set the destination indexes of input frames.
18339 This is space or '|' separated list of indexes that maps input frames to output
18340 frames. Number of indexes also sets maximal value that each index may have.
18341 '-1' index have special meaning and that is to drop frame.
18344 The first frame has the index 0. The default is to keep the input unchanged.
18346 @subsection Examples
18350 Swap second and third frame of every three frames of the input:
18352 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
18356 Swap 10th and 1st frame of every ten frames of the input:
18358 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
18362 @section shufflepixels
18364 Reorder pixels in video frames.
18366 This filter accepts the following options:
18370 Set shuffle direction. Can be forward or inverse direction.
18371 Default direction is forward.
18374 Set shuffle mode. Can be horizontal, vertical or block mode.
18378 Set shuffle block_size. In case of horizontal shuffle mode only width
18379 part of size is used, and in case of vertical shuffle mode only height
18380 part of size is used.
18383 Set random seed used with shuffling pixels. Mainly useful to set to be able
18384 to reverse filtering process to get original input.
18385 For example, to reverse forward shuffle you need to use same parameters
18386 and exact same seed and to set direction to inverse.
18389 @section shuffleplanes
18391 Reorder and/or duplicate video planes.
18393 It accepts the following parameters:
18398 The index of the input plane to be used as the first output plane.
18401 The index of the input plane to be used as the second output plane.
18404 The index of the input plane to be used as the third output plane.
18407 The index of the input plane to be used as the fourth output plane.
18411 The first plane has the index 0. The default is to keep the input unchanged.
18413 @subsection Examples
18417 Swap the second and third planes of the input:
18419 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
18423 @anchor{signalstats}
18424 @section signalstats
18425 Evaluate various visual metrics that assist in determining issues associated
18426 with the digitization of analog video media.
18428 By default the filter will log these metadata values:
18432 Display the minimal Y value contained within the input frame. Expressed in
18436 Display the Y value at the 10% percentile within the input frame. Expressed in
18440 Display the average Y value within the input frame. Expressed in range of
18444 Display the Y value at the 90% percentile within the input frame. Expressed in
18448 Display the maximum Y value contained within the input frame. Expressed in
18452 Display the minimal U value contained within the input frame. Expressed in
18456 Display the U value at the 10% percentile within the input frame. Expressed in
18460 Display the average U value within the input frame. Expressed in range of
18464 Display the U value at the 90% percentile within the input frame. Expressed in
18468 Display the maximum U value contained within the input frame. Expressed in
18472 Display the minimal V value contained within the input frame. Expressed in
18476 Display the V value at the 10% percentile within the input frame. Expressed in
18480 Display the average V value within the input frame. Expressed in range of
18484 Display the V value at the 90% percentile within the input frame. Expressed in
18488 Display the maximum V value contained within the input frame. Expressed in
18492 Display the minimal saturation value contained within the input frame.
18493 Expressed in range of [0-~181.02].
18496 Display the saturation value at the 10% percentile within the input frame.
18497 Expressed in range of [0-~181.02].
18500 Display the average saturation value within the input frame. Expressed in range
18504 Display the saturation value at the 90% percentile within the input frame.
18505 Expressed in range of [0-~181.02].
18508 Display the maximum saturation value contained within the input frame.
18509 Expressed in range of [0-~181.02].
18512 Display the median value for hue within the input frame. Expressed in range of
18516 Display the average value for hue within the input frame. Expressed in range of
18520 Display the average of sample value difference between all values of the Y
18521 plane in the current frame and corresponding values of the previous input frame.
18522 Expressed in range of [0-255].
18525 Display the average of sample value difference between all values of the U
18526 plane in the current frame and corresponding values of the previous input frame.
18527 Expressed in range of [0-255].
18530 Display the average of sample value difference between all values of the V
18531 plane in the current frame and corresponding values of the previous input frame.
18532 Expressed in range of [0-255].
18535 Display bit depth of Y plane in current frame.
18536 Expressed in range of [0-16].
18539 Display bit depth of U plane in current frame.
18540 Expressed in range of [0-16].
18543 Display bit depth of V plane in current frame.
18544 Expressed in range of [0-16].
18547 The filter accepts the following options:
18553 @option{stat} specify an additional form of image analysis.
18554 @option{out} output video with the specified type of pixel highlighted.
18556 Both options accept the following values:
18560 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
18561 unlike the neighboring pixels of the same field. Examples of temporal outliers
18562 include the results of video dropouts, head clogs, or tape tracking issues.
18565 Identify @var{vertical line repetition}. Vertical line repetition includes
18566 similar rows of pixels within a frame. In born-digital video vertical line
18567 repetition is common, but this pattern is uncommon in video digitized from an
18568 analog source. When it occurs in video that results from the digitization of an
18569 analog source it can indicate concealment from a dropout compensator.
18572 Identify pixels that fall outside of legal broadcast range.
18576 Set the highlight color for the @option{out} option. The default color is
18580 @subsection Examples
18584 Output data of various video metrics:
18586 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
18590 Output specific data about the minimum and maximum values of the Y plane per frame:
18592 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
18596 Playback video while highlighting pixels that are outside of broadcast range in red.
18598 ffplay example.mov -vf signalstats="out=brng:color=red"
18602 Playback video with signalstats metadata drawn over the frame.
18604 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
18607 The contents of signalstat_drawtext.txt used in the command are:
18610 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
18611 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
18612 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
18613 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
18621 Calculates the MPEG-7 Video Signature. The filter can handle more than one
18622 input. In this case the matching between the inputs can be calculated additionally.
18623 The filter always passes through the first input. The signature of each stream can
18624 be written into a file.
18626 It accepts the following options:
18630 Enable or disable the matching process.
18632 Available values are:
18636 Disable the calculation of a matching (default).
18638 Calculate the matching for the whole video and output whether the whole video
18639 matches or only parts.
18641 Calculate only until a matching is found or the video ends. Should be faster in
18646 Set the number of inputs. The option value must be a non negative integer.
18647 Default value is 1.
18650 Set the path to which the output is written. If there is more than one input,
18651 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
18652 integer), that will be replaced with the input number. If no filename is
18653 specified, no output will be written. This is the default.
18656 Choose the output format.
18658 Available values are:
18662 Use the specified binary representation (default).
18664 Use the specified xml representation.
18668 Set threshold to detect one word as similar. The option value must be an integer
18669 greater than zero. The default value is 9000.
18672 Set threshold to detect all words as similar. The option value must be an integer
18673 greater than zero. The default value is 60000.
18676 Set threshold to detect frames as similar. The option value must be an integer
18677 greater than zero. The default value is 116.
18680 Set the minimum length of a sequence in frames to recognize it as matching
18681 sequence. The option value must be a non negative integer value.
18682 The default value is 0.
18685 Set the minimum relation, that matching frames to all frames must have.
18686 The option value must be a double value between 0 and 1. The default value is 0.5.
18689 @subsection Examples
18693 To calculate the signature of an input video and store it in signature.bin:
18695 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
18699 To detect whether two videos match and store the signatures in XML format in
18700 signature0.xml and signature1.xml:
18702 ffmpeg -i input1.mkv -i input2.mkv -filter_complex "[0:v][1:v] signature=nb_inputs=2:detectmode=full:format=xml:filename=signature%d.xml" -map :v -f null -
18710 Blur the input video without impacting the outlines.
18712 It accepts the following options:
18715 @item luma_radius, lr
18716 Set the luma radius. The option value must be a float number in
18717 the range [0.1,5.0] that specifies the variance of the gaussian filter
18718 used to blur the image (slower if larger). Default value is 1.0.
18720 @item luma_strength, ls
18721 Set the luma strength. The option value must be a float number
18722 in the range [-1.0,1.0] that configures the blurring. A value included
18723 in [0.0,1.0] will blur the image whereas a value included in
18724 [-1.0,0.0] will sharpen the image. Default value is 1.0.
18726 @item luma_threshold, lt
18727 Set the luma threshold used as a coefficient to determine
18728 whether a pixel should be blurred or not. The option value must be an
18729 integer in the range [-30,30]. A value of 0 will filter all the image,
18730 a value included in [0,30] will filter flat areas and a value included
18731 in [-30,0] will filter edges. Default value is 0.
18733 @item chroma_radius, cr
18734 Set the chroma radius. The option value must be a float number in
18735 the range [0.1,5.0] that specifies the variance of the gaussian filter
18736 used to blur the image (slower if larger). Default value is @option{luma_radius}.
18738 @item chroma_strength, cs
18739 Set the chroma strength. The option value must be a float number
18740 in the range [-1.0,1.0] that configures the blurring. A value included
18741 in [0.0,1.0] will blur the image whereas a value included in
18742 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
18744 @item chroma_threshold, ct
18745 Set the chroma threshold used as a coefficient to determine
18746 whether a pixel should be blurred or not. The option value must be an
18747 integer in the range [-30,30]. A value of 0 will filter all the image,
18748 a value included in [0,30] will filter flat areas and a value included
18749 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
18752 If a chroma option is not explicitly set, the corresponding luma value
18756 Apply sobel operator to input video stream.
18758 The filter accepts the following option:
18762 Set which planes will be processed, unprocessed planes will be copied.
18763 By default value 0xf, all planes will be processed.
18766 Set value which will be multiplied with filtered result.
18769 Set value which will be added to filtered result.
18772 @subsection Commands
18774 This filter supports the all above options as @ref{commands}.
18779 Apply a simple postprocessing filter that compresses and decompresses the image
18780 at several (or - in the case of @option{quality} level @code{6} - all) shifts
18781 and average the results.
18783 The filter accepts the following options:
18787 Set quality. This option defines the number of levels for averaging. It accepts
18788 an integer in the range 0-6. If set to @code{0}, the filter will have no
18789 effect. A value of @code{6} means the higher quality. For each increment of
18790 that value the speed drops by a factor of approximately 2. Default value is
18794 Force a constant quantization parameter. If not set, the filter will use the QP
18795 from the video stream (if available).
18798 Set thresholding mode. Available modes are:
18802 Set hard thresholding (default).
18804 Set soft thresholding (better de-ringing effect, but likely blurrier).
18807 @item use_bframe_qp
18808 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
18809 option may cause flicker since the B-Frames have often larger QP. Default is
18810 @code{0} (not enabled).
18813 @subsection Commands
18815 This filter supports the following commands:
18817 @item quality, level
18818 Set quality level. The value @code{max} can be used to set the maximum level,
18819 currently @code{6}.
18825 Scale the input by applying one of the super-resolution methods based on
18826 convolutional neural networks. Supported models:
18830 Super-Resolution Convolutional Neural Network model (SRCNN).
18831 See @url{https://arxiv.org/abs/1501.00092}.
18834 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
18835 See @url{https://arxiv.org/abs/1609.05158}.
18838 Training scripts as well as scripts for model file (.pb) saving can be found at
18839 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
18840 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
18842 Native model files (.model) can be generated from TensorFlow model
18843 files (.pb) by using tools/python/convert.py
18845 The filter accepts the following options:
18849 Specify which DNN backend to use for model loading and execution. This option accepts
18850 the following values:
18854 Native implementation of DNN loading and execution.
18857 TensorFlow backend. To enable this backend you
18858 need to install the TensorFlow for C library (see
18859 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
18860 @code{--enable-libtensorflow}
18863 Default value is @samp{native}.
18866 Set path to model file specifying network architecture and its parameters.
18867 Note that different backends use different file formats. TensorFlow backend
18868 can load files for both formats, while native backend can load files for only
18872 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
18873 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
18874 input upscaled using bicubic upscaling with proper scale factor.
18877 This feature can also be finished with @ref{dnn_processing} filter.
18881 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
18883 This filter takes in input two input videos, the first input is
18884 considered the "main" source and is passed unchanged to the
18885 output. The second input is used as a "reference" video for computing
18888 Both video inputs must have the same resolution and pixel format for
18889 this filter to work correctly. Also it assumes that both inputs
18890 have the same number of frames, which are compared one by one.
18892 The filter stores the calculated SSIM of each frame.
18894 The description of the accepted parameters follows.
18897 @item stats_file, f
18898 If specified the filter will use the named file to save the SSIM of
18899 each individual frame. When filename equals "-" the data is sent to
18903 The file printed if @var{stats_file} is selected, contains a sequence of
18904 key/value pairs of the form @var{key}:@var{value} for each compared
18907 A description of each shown parameter follows:
18911 sequential number of the input frame, starting from 1
18913 @item Y, U, V, R, G, B
18914 SSIM of the compared frames for the component specified by the suffix.
18917 SSIM of the compared frames for the whole frame.
18920 Same as above but in dB representation.
18923 This filter also supports the @ref{framesync} options.
18925 @subsection Examples
18930 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
18931 [main][ref] ssim="stats_file=stats.log" [out]
18934 On this example the input file being processed is compared with the
18935 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
18936 is stored in @file{stats.log}.
18939 Another example with both psnr and ssim at same time:
18941 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
18945 Another example with different containers:
18947 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]ssim" -f null -
18953 Convert between different stereoscopic image formats.
18955 The filters accept the following options:
18959 Set stereoscopic image format of input.
18961 Available values for input image formats are:
18964 side by side parallel (left eye left, right eye right)
18967 side by side crosseye (right eye left, left eye right)
18970 side by side parallel with half width resolution
18971 (left eye left, right eye right)
18974 side by side crosseye with half width resolution
18975 (right eye left, left eye right)
18979 above-below (left eye above, right eye below)
18983 above-below (right eye above, left eye below)
18987 above-below with half height resolution
18988 (left eye above, right eye below)
18992 above-below with half height resolution
18993 (right eye above, left eye below)
18996 alternating frames (left eye first, right eye second)
18999 alternating frames (right eye first, left eye second)
19002 interleaved rows (left eye has top row, right eye starts on next row)
19005 interleaved rows (right eye has top row, left eye starts on next row)
19008 interleaved columns, left eye first
19011 interleaved columns, right eye first
19013 Default value is @samp{sbsl}.
19017 Set stereoscopic image format of output.
19021 side by side parallel (left eye left, right eye right)
19024 side by side crosseye (right eye left, left eye right)
19027 side by side parallel with half width resolution
19028 (left eye left, right eye right)
19031 side by side crosseye with half width resolution
19032 (right eye left, left eye right)
19036 above-below (left eye above, right eye below)
19040 above-below (right eye above, left eye below)
19044 above-below with half height resolution
19045 (left eye above, right eye below)
19049 above-below with half height resolution
19050 (right eye above, left eye below)
19053 alternating frames (left eye first, right eye second)
19056 alternating frames (right eye first, left eye second)
19059 interleaved rows (left eye has top row, right eye starts on next row)
19062 interleaved rows (right eye has top row, left eye starts on next row)
19065 anaglyph red/blue gray
19066 (red filter on left eye, blue filter on right eye)
19069 anaglyph red/green gray
19070 (red filter on left eye, green filter on right eye)
19073 anaglyph red/cyan gray
19074 (red filter on left eye, cyan filter on right eye)
19077 anaglyph red/cyan half colored
19078 (red filter on left eye, cyan filter on right eye)
19081 anaglyph red/cyan color
19082 (red filter on left eye, cyan filter on right eye)
19085 anaglyph red/cyan color optimized with the least squares projection of dubois
19086 (red filter on left eye, cyan filter on right eye)
19089 anaglyph green/magenta gray
19090 (green filter on left eye, magenta filter on right eye)
19093 anaglyph green/magenta half colored
19094 (green filter on left eye, magenta filter on right eye)
19097 anaglyph green/magenta colored
19098 (green filter on left eye, magenta filter on right eye)
19101 anaglyph green/magenta color optimized with the least squares projection of dubois
19102 (green filter on left eye, magenta filter on right eye)
19105 anaglyph yellow/blue gray
19106 (yellow filter on left eye, blue filter on right eye)
19109 anaglyph yellow/blue half colored
19110 (yellow filter on left eye, blue filter on right eye)
19113 anaglyph yellow/blue colored
19114 (yellow filter on left eye, blue filter on right eye)
19117 anaglyph yellow/blue color optimized with the least squares projection of dubois
19118 (yellow filter on left eye, blue filter on right eye)
19121 mono output (left eye only)
19124 mono output (right eye only)
19127 checkerboard, left eye first
19130 checkerboard, right eye first
19133 interleaved columns, left eye first
19136 interleaved columns, right eye first
19142 Default value is @samp{arcd}.
19145 @subsection Examples
19149 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
19155 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
19161 @section streamselect, astreamselect
19162 Select video or audio streams.
19164 The filter accepts the following options:
19168 Set number of inputs. Default is 2.
19171 Set input indexes to remap to outputs.
19174 @subsection Commands
19176 The @code{streamselect} and @code{astreamselect} filter supports the following
19181 Set input indexes to remap to outputs.
19184 @subsection Examples
19188 Select first 5 seconds 1st stream and rest of time 2nd stream:
19190 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
19194 Same as above, but for audio:
19196 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
19203 Draw subtitles on top of input video using the libass library.
19205 To enable compilation of this filter you need to configure FFmpeg with
19206 @code{--enable-libass}. This filter also requires a build with libavcodec and
19207 libavformat to convert the passed subtitles file to ASS (Advanced Substation
19208 Alpha) subtitles format.
19210 The filter accepts the following options:
19214 Set the filename of the subtitle file to read. It must be specified.
19216 @item original_size
19217 Specify the size of the original video, the video for which the ASS file
19218 was composed. For the syntax of this option, check the
19219 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19220 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
19221 correctly scale the fonts if the aspect ratio has been changed.
19224 Set a directory path containing fonts that can be used by the filter.
19225 These fonts will be used in addition to whatever the font provider uses.
19228 Process alpha channel, by default alpha channel is untouched.
19231 Set subtitles input character encoding. @code{subtitles} filter only. Only
19232 useful if not UTF-8.
19234 @item stream_index, si
19235 Set subtitles stream index. @code{subtitles} filter only.
19238 Override default style or script info parameters of the subtitles. It accepts a
19239 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
19242 If the first key is not specified, it is assumed that the first value
19243 specifies the @option{filename}.
19245 For example, to render the file @file{sub.srt} on top of the input
19246 video, use the command:
19251 which is equivalent to:
19253 subtitles=filename=sub.srt
19256 To render the default subtitles stream from file @file{video.mkv}, use:
19258 subtitles=video.mkv
19261 To render the second subtitles stream from that file, use:
19263 subtitles=video.mkv:si=1
19266 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
19267 @code{DejaVu Serif}, use:
19269 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
19272 @section super2xsai
19274 Scale the input by 2x and smooth using the Super2xSaI (Scale and
19275 Interpolate) pixel art scaling algorithm.
19277 Useful for enlarging pixel art images without reducing sharpness.
19281 Swap two rectangular objects in video.
19283 This filter accepts the following options:
19293 Set 1st rect x coordinate.
19296 Set 1st rect y coordinate.
19299 Set 2nd rect x coordinate.
19302 Set 2nd rect y coordinate.
19304 All expressions are evaluated once for each frame.
19307 The all options are expressions containing the following constants:
19312 The input width and height.
19315 same as @var{w} / @var{h}
19318 input sample aspect ratio
19321 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
19324 The number of the input frame, starting from 0.
19327 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
19330 the position in the file of the input frame, NAN if unknown
19333 @subsection Commands
19335 This filter supports the all above options as @ref{commands}.
19341 Blend successive video frames.
19347 Apply telecine process to the video.
19349 This filter accepts the following options:
19358 The default value is @code{top}.
19362 A string of numbers representing the pulldown pattern you wish to apply.
19363 The default value is @code{23}.
19367 Some typical patterns:
19372 24p: 2332 (preferred)
19379 24p: 222222222223 ("Euro pulldown")
19384 @section thistogram
19386 Compute and draw a color distribution histogram for the input video across time.
19388 Unlike @ref{histogram} video filter which only shows histogram of single input frame
19389 at certain time, this filter shows also past histograms of number of frames defined
19390 by @code{width} option.
19392 The computed histogram is a representation of the color component
19393 distribution in an image.
19395 The filter accepts the following options:
19399 Set width of single color component output. Default value is @code{0}.
19400 Value of @code{0} means width will be picked from input video.
19401 This also set number of passed histograms to keep.
19402 Allowed range is [0, 8192].
19404 @item display_mode, d
19406 It accepts the following values:
19409 Per color component graphs are placed below each other.
19412 Per color component graphs are placed side by side.
19415 Presents information identical to that in the @code{parade}, except
19416 that the graphs representing color components are superimposed directly
19419 Default is @code{stack}.
19421 @item levels_mode, m
19422 Set mode. Can be either @code{linear}, or @code{logarithmic}.
19423 Default is @code{linear}.
19425 @item components, c
19426 Set what color components to display.
19427 Default is @code{7}.
19430 Set background opacity. Default is @code{0.9}.
19433 Show envelope. Default is disabled.
19436 Set envelope color. Default is @code{gold}.
19441 Available values for slide is:
19444 Draw new frame when right border is reached.
19447 Replace old columns with new ones.
19450 Scroll from right to left.
19453 Scroll from left to right.
19456 Draw single picture.
19459 Default is @code{replace}.
19464 Apply threshold effect to video stream.
19466 This filter needs four video streams to perform thresholding.
19467 First stream is stream we are filtering.
19468 Second stream is holding threshold values, third stream is holding min values,
19469 and last, fourth stream is holding max values.
19471 The filter accepts the following option:
19475 Set which planes will be processed, unprocessed planes will be copied.
19476 By default value 0xf, all planes will be processed.
19479 For example if first stream pixel's component value is less then threshold value
19480 of pixel component from 2nd threshold stream, third stream value will picked,
19481 otherwise fourth stream pixel component value will be picked.
19483 Using color source filter one can perform various types of thresholding:
19485 @subsection Examples
19489 Binary threshold, using gray color as threshold:
19491 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
19495 Inverted binary threshold, using gray color as threshold:
19497 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
19501 Truncate binary threshold, using gray color as threshold:
19503 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
19507 Threshold to zero, using gray color as threshold:
19509 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
19513 Inverted threshold to zero, using gray color as threshold:
19515 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
19520 Select the most representative frame in a given sequence of consecutive frames.
19522 The filter accepts the following options:
19526 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
19527 will pick one of them, and then handle the next batch of @var{n} frames until
19528 the end. Default is @code{100}.
19531 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
19532 value will result in a higher memory usage, so a high value is not recommended.
19534 @subsection Examples
19538 Extract one picture each 50 frames:
19544 Complete example of a thumbnail creation with @command{ffmpeg}:
19546 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
19553 Tile several successive frames together.
19555 The @ref{untile} filter can do the reverse.
19557 The filter accepts the following options:
19562 Set the grid size (i.e. the number of lines and columns). For the syntax of
19563 this option, check the
19564 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19567 Set the maximum number of frames to render in the given area. It must be less
19568 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
19569 the area will be used.
19572 Set the outer border margin in pixels.
19575 Set the inner border thickness (i.e. the number of pixels between frames). For
19576 more advanced padding options (such as having different values for the edges),
19577 refer to the pad video filter.
19580 Specify the color of the unused area. For the syntax of this option, check the
19581 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19582 The default value of @var{color} is "black".
19585 Set the number of frames to overlap when tiling several successive frames together.
19586 The value must be between @code{0} and @var{nb_frames - 1}.
19589 Set the number of frames to initially be empty before displaying first output frame.
19590 This controls how soon will one get first output frame.
19591 The value must be between @code{0} and @var{nb_frames - 1}.
19594 @subsection Examples
19598 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
19600 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
19602 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
19603 duplicating each output frame to accommodate the originally detected frame
19607 Display @code{5} pictures in an area of @code{3x2} frames,
19608 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
19609 mixed flat and named options:
19611 tile=3x2:nb_frames=5:padding=7:margin=2
19615 @section tinterlace
19617 Perform various types of temporal field interlacing.
19619 Frames are counted starting from 1, so the first input frame is
19622 The filter accepts the following options:
19627 Specify the mode of the interlacing. This option can also be specified
19628 as a value alone. See below for a list of values for this option.
19630 Available values are:
19634 Move odd frames into the upper field, even into the lower field,
19635 generating a double height frame at half frame rate.
19639 Frame 1 Frame 2 Frame 3 Frame 4
19641 11111 22222 33333 44444
19642 11111 22222 33333 44444
19643 11111 22222 33333 44444
19644 11111 22222 33333 44444
19658 Only output odd frames, even frames are dropped, generating a frame with
19659 unchanged height at half frame rate.
19664 Frame 1 Frame 2 Frame 3 Frame 4
19666 11111 22222 33333 44444
19667 11111 22222 33333 44444
19668 11111 22222 33333 44444
19669 11111 22222 33333 44444
19679 Only output even frames, odd frames are dropped, generating a frame with
19680 unchanged height at half frame rate.
19685 Frame 1 Frame 2 Frame 3 Frame 4
19687 11111 22222 33333 44444
19688 11111 22222 33333 44444
19689 11111 22222 33333 44444
19690 11111 22222 33333 44444
19700 Expand each frame to full height, but pad alternate lines with black,
19701 generating a frame with double height at the same input frame rate.
19706 Frame 1 Frame 2 Frame 3 Frame 4
19708 11111 22222 33333 44444
19709 11111 22222 33333 44444
19710 11111 22222 33333 44444
19711 11111 22222 33333 44444
19714 11111 ..... 33333 .....
19715 ..... 22222 ..... 44444
19716 11111 ..... 33333 .....
19717 ..... 22222 ..... 44444
19718 11111 ..... 33333 .....
19719 ..... 22222 ..... 44444
19720 11111 ..... 33333 .....
19721 ..... 22222 ..... 44444
19725 @item interleave_top, 4
19726 Interleave the upper field from odd frames with the lower field from
19727 even frames, generating a frame with unchanged height at half frame rate.
19732 Frame 1 Frame 2 Frame 3 Frame 4
19734 11111<- 22222 33333<- 44444
19735 11111 22222<- 33333 44444<-
19736 11111<- 22222 33333<- 44444
19737 11111 22222<- 33333 44444<-
19747 @item interleave_bottom, 5
19748 Interleave the lower field from odd frames with the upper field from
19749 even frames, generating a frame with unchanged height at half frame rate.
19754 Frame 1 Frame 2 Frame 3 Frame 4
19756 11111 22222<- 33333 44444<-
19757 11111<- 22222 33333<- 44444
19758 11111 22222<- 33333 44444<-
19759 11111<- 22222 33333<- 44444
19769 @item interlacex2, 6
19770 Double frame rate with unchanged height. Frames are inserted each
19771 containing the second temporal field from the previous input frame and
19772 the first temporal field from the next input frame. This mode relies on
19773 the top_field_first flag. Useful for interlaced video displays with no
19774 field synchronisation.
19779 Frame 1 Frame 2 Frame 3 Frame 4
19781 11111 22222 33333 44444
19782 11111 22222 33333 44444
19783 11111 22222 33333 44444
19784 11111 22222 33333 44444
19787 11111 22222 22222 33333 33333 44444 44444
19788 11111 11111 22222 22222 33333 33333 44444
19789 11111 22222 22222 33333 33333 44444 44444
19790 11111 11111 22222 22222 33333 33333 44444
19795 Move odd frames into the upper field, even into the lower field,
19796 generating a double height frame at same frame rate.
19801 Frame 1 Frame 2 Frame 3 Frame 4
19803 11111 22222 33333 44444
19804 11111 22222 33333 44444
19805 11111 22222 33333 44444
19806 11111 22222 33333 44444
19809 11111 33333 33333 55555
19810 22222 22222 44444 44444
19811 11111 33333 33333 55555
19812 22222 22222 44444 44444
19813 11111 33333 33333 55555
19814 22222 22222 44444 44444
19815 11111 33333 33333 55555
19816 22222 22222 44444 44444
19821 Numeric values are deprecated but are accepted for backward
19822 compatibility reasons.
19824 Default mode is @code{merge}.
19827 Specify flags influencing the filter process.
19829 Available value for @var{flags} is:
19832 @item low_pass_filter, vlpf
19833 Enable linear vertical low-pass filtering in the filter.
19834 Vertical low-pass filtering is required when creating an interlaced
19835 destination from a progressive source which contains high-frequency
19836 vertical detail. Filtering will reduce interlace 'twitter' and Moire
19839 @item complex_filter, cvlpf
19840 Enable complex vertical low-pass filtering.
19841 This will slightly less reduce interlace 'twitter' and Moire
19842 patterning but better retain detail and subjective sharpness impression.
19845 Bypass already interlaced frames, only adjust the frame rate.
19848 Vertical low-pass filtering and bypassing already interlaced frames can only be
19849 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
19854 Pick median pixels from several successive input video frames.
19856 The filter accepts the following options:
19860 Set radius of median filter.
19861 Default is 1. Allowed range is from 1 to 127.
19864 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
19867 Set median percentile. Default value is @code{0.5}.
19868 Default value of @code{0.5} will pick always median values, while @code{0} will pick
19869 minimum values, and @code{1} maximum values.
19872 @subsection Commands
19874 This filter supports all above options as @ref{commands}, excluding option @code{radius}.
19876 @section tmidequalizer
19878 Apply Temporal Midway Video Equalization effect.
19880 Midway Video Equalization adjusts a sequence of video frames to have the same
19881 histograms, while maintaining their dynamics as much as possible. It's
19882 useful for e.g. matching exposures from a video frames sequence.
19884 This filter accepts the following option:
19888 Set filtering radius. Default is @code{5}. Allowed range is from 1 to 127.
19891 Set filtering sigma. Default is @code{0.5}. This controls strength of filtering.
19892 Setting this option to 0 effectively does nothing.
19895 Set which planes to process. Default is @code{15}, which is all available planes.
19900 Mix successive video frames.
19902 A description of the accepted options follows.
19906 The number of successive frames to mix. If unspecified, it defaults to 3.
19909 Specify weight of each input video frame.
19910 Each weight is separated by space. If number of weights is smaller than
19911 number of @var{frames} last specified weight will be used for all remaining
19915 Specify scale, if it is set it will be multiplied with sum
19916 of each weight multiplied with pixel values to give final destination
19917 pixel value. By default @var{scale} is auto scaled to sum of weights.
19920 @subsection Examples
19924 Average 7 successive frames:
19926 tmix=frames=7:weights="1 1 1 1 1 1 1"
19930 Apply simple temporal convolution:
19932 tmix=frames=3:weights="-1 3 -1"
19936 Similar as above but only showing temporal differences:
19938 tmix=frames=3:weights="-1 2 -1":scale=1
19942 @subsection Commands
19944 This filter supports the following commands:
19948 Syntax is same as option with same name.
19953 Tone map colors from different dynamic ranges.
19955 This filter expects data in single precision floating point, as it needs to
19956 operate on (and can output) out-of-range values. Another filter, such as
19957 @ref{zscale}, is needed to convert the resulting frame to a usable format.
19959 The tonemapping algorithms implemented only work on linear light, so input
19960 data should be linearized beforehand (and possibly correctly tagged).
19963 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
19966 @subsection Options
19967 The filter accepts the following options.
19971 Set the tone map algorithm to use.
19973 Possible values are:
19976 Do not apply any tone map, only desaturate overbright pixels.
19979 Hard-clip any out-of-range values. Use it for perfect color accuracy for
19980 in-range values, while distorting out-of-range values.
19983 Stretch the entire reference gamut to a linear multiple of the display.
19986 Fit a logarithmic transfer between the tone curves.
19989 Preserve overall image brightness with a simple curve, using nonlinear
19990 contrast, which results in flattening details and degrading color accuracy.
19993 Preserve both dark and bright details better than @var{reinhard}, at the cost
19994 of slightly darkening everything. Use it when detail preservation is more
19995 important than color and brightness accuracy.
19998 Smoothly map out-of-range values, while retaining contrast and colors for
19999 in-range material as much as possible. Use it when color accuracy is more
20000 important than detail preservation.
20006 Tune the tone mapping algorithm.
20008 This affects the following algorithms:
20014 Specifies the scale factor to use while stretching.
20018 Specifies the exponent of the function.
20022 Specify an extra linear coefficient to multiply into the signal before clipping.
20026 Specify the local contrast coefficient at the display peak.
20027 Default to 0.5, which means that in-gamut values will be about half as bright
20034 Specify the transition point from linear to mobius transform. Every value
20035 below this point is guaranteed to be mapped 1:1. The higher the value, the
20036 more accurate the result will be, at the cost of losing bright details.
20037 Default to 0.3, which due to the steep initial slope still preserves in-range
20038 colors fairly accurately.
20042 Apply desaturation for highlights that exceed this level of brightness. The
20043 higher the parameter, the more color information will be preserved. This
20044 setting helps prevent unnaturally blown-out colors for super-highlights, by
20045 (smoothly) turning into white instead. This makes images feel more natural,
20046 at the cost of reducing information about out-of-range colors.
20048 The default of 2.0 is somewhat conservative and will mostly just apply to
20049 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
20051 This option works only if the input frame has a supported color tag.
20054 Override signal/nominal/reference peak with this value. Useful when the
20055 embedded peak information in display metadata is not reliable or when tone
20056 mapping from a lower range to a higher range.
20061 Temporarily pad video frames.
20063 The filter accepts the following options:
20067 Specify number of delay frames before input video stream. Default is 0.
20070 Specify number of padding frames after input video stream.
20071 Set to -1 to pad indefinitely. Default is 0.
20074 Set kind of frames added to beginning of stream.
20075 Can be either @var{add} or @var{clone}.
20076 With @var{add} frames of solid-color are added.
20077 With @var{clone} frames are clones of first frame.
20078 Default is @var{add}.
20081 Set kind of frames added to end of stream.
20082 Can be either @var{add} or @var{clone}.
20083 With @var{add} frames of solid-color are added.
20084 With @var{clone} frames are clones of last frame.
20085 Default is @var{add}.
20087 @item start_duration, stop_duration
20088 Specify the duration of the start/stop delay. See
20089 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20090 for the accepted syntax.
20091 These options override @var{start} and @var{stop}. Default is 0.
20094 Specify the color of the padded area. For the syntax of this option,
20095 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
20096 manual,ffmpeg-utils}.
20098 The default value of @var{color} is "black".
20104 Transpose rows with columns in the input video and optionally flip it.
20106 It accepts the following parameters:
20111 Specify the transposition direction.
20113 Can assume the following values:
20115 @item 0, 4, cclock_flip
20116 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
20124 Rotate by 90 degrees clockwise, that is:
20132 Rotate by 90 degrees counterclockwise, that is:
20139 @item 3, 7, clock_flip
20140 Rotate by 90 degrees clockwise and vertically flip, that is:
20148 For values between 4-7, the transposition is only done if the input
20149 video geometry is portrait and not landscape. These values are
20150 deprecated, the @code{passthrough} option should be used instead.
20152 Numerical values are deprecated, and should be dropped in favor of
20153 symbolic constants.
20156 Do not apply the transposition if the input geometry matches the one
20157 specified by the specified value. It accepts the following values:
20160 Always apply transposition.
20162 Preserve portrait geometry (when @var{height} >= @var{width}).
20164 Preserve landscape geometry (when @var{width} >= @var{height}).
20167 Default value is @code{none}.
20170 For example to rotate by 90 degrees clockwise and preserve portrait
20173 transpose=dir=1:passthrough=portrait
20176 The command above can also be specified as:
20178 transpose=1:portrait
20181 @section transpose_npp
20183 Transpose rows with columns in the input video and optionally flip it.
20184 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
20186 It accepts the following parameters:
20191 Specify the transposition direction.
20193 Can assume the following values:
20196 Rotate by 90 degrees counterclockwise and vertically flip. (default)
20199 Rotate by 90 degrees clockwise.
20202 Rotate by 90 degrees counterclockwise.
20205 Rotate by 90 degrees clockwise and vertically flip.
20209 Do not apply the transposition if the input geometry matches the one
20210 specified by the specified value. It accepts the following values:
20213 Always apply transposition. (default)
20215 Preserve portrait geometry (when @var{height} >= @var{width}).
20217 Preserve landscape geometry (when @var{width} >= @var{height}).
20223 Trim the input so that the output contains one continuous subpart of the input.
20225 It accepts the following parameters:
20228 Specify the time of the start of the kept section, i.e. the frame with the
20229 timestamp @var{start} will be the first frame in the output.
20232 Specify the time of the first frame that will be dropped, i.e. the frame
20233 immediately preceding the one with the timestamp @var{end} will be the last
20234 frame in the output.
20237 This is the same as @var{start}, except this option sets the start timestamp
20238 in timebase units instead of seconds.
20241 This is the same as @var{end}, except this option sets the end timestamp
20242 in timebase units instead of seconds.
20245 The maximum duration of the output in seconds.
20248 The number of the first frame that should be passed to the output.
20251 The number of the first frame that should be dropped.
20254 @option{start}, @option{end}, and @option{duration} are expressed as time
20255 duration specifications; see
20256 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20257 for the accepted syntax.
20259 Note that the first two sets of the start/end options and the @option{duration}
20260 option look at the frame timestamp, while the _frame variants simply count the
20261 frames that pass through the filter. Also note that this filter does not modify
20262 the timestamps. If you wish for the output timestamps to start at zero, insert a
20263 setpts filter after the trim filter.
20265 If multiple start or end options are set, this filter tries to be greedy and
20266 keep all the frames that match at least one of the specified constraints. To keep
20267 only the part that matches all the constraints at once, chain multiple trim
20270 The defaults are such that all the input is kept. So it is possible to set e.g.
20271 just the end values to keep everything before the specified time.
20276 Drop everything except the second minute of input:
20278 ffmpeg -i INPUT -vf trim=60:120
20282 Keep only the first second:
20284 ffmpeg -i INPUT -vf trim=duration=1
20289 @section unpremultiply
20290 Apply alpha unpremultiply effect to input video stream using first plane
20291 of second stream as alpha.
20293 Both streams must have same dimensions and same pixel format.
20295 The filter accepts the following option:
20299 Set which planes will be processed, unprocessed planes will be copied.
20300 By default value 0xf, all planes will be processed.
20302 If the format has 1 or 2 components, then luma is bit 0.
20303 If the format has 3 or 4 components:
20304 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
20305 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
20306 If present, the alpha channel is always the last bit.
20309 Do not require 2nd input for processing, instead use alpha plane from input stream.
20315 Sharpen or blur the input video.
20317 It accepts the following parameters:
20320 @item luma_msize_x, lx
20321 Set the luma matrix horizontal size. It must be an odd integer between
20322 3 and 23. The default value is 5.
20324 @item luma_msize_y, ly
20325 Set the luma matrix vertical size. It must be an odd integer between 3
20326 and 23. The default value is 5.
20328 @item luma_amount, la
20329 Set the luma effect strength. It must be a floating point number, reasonable
20330 values lay between -1.5 and 1.5.
20332 Negative values will blur the input video, while positive values will
20333 sharpen it, a value of zero will disable the effect.
20335 Default value is 1.0.
20337 @item chroma_msize_x, cx
20338 Set the chroma matrix horizontal size. It must be an odd integer
20339 between 3 and 23. The default value is 5.
20341 @item chroma_msize_y, cy
20342 Set the chroma matrix vertical size. It must be an odd integer
20343 between 3 and 23. The default value is 5.
20345 @item chroma_amount, ca
20346 Set the chroma effect strength. It must be a floating point number, reasonable
20347 values lay between -1.5 and 1.5.
20349 Negative values will blur the input video, while positive values will
20350 sharpen it, a value of zero will disable the effect.
20352 Default value is 0.0.
20356 All parameters are optional and default to the equivalent of the
20357 string '5:5:1.0:5:5:0.0'.
20359 @subsection Examples
20363 Apply strong luma sharpen effect:
20365 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
20369 Apply a strong blur of both luma and chroma parameters:
20371 unsharp=7:7:-2:7:7:-2
20378 Decompose a video made of tiled images into the individual images.
20380 The frame rate of the output video is the frame rate of the input video
20381 multiplied by the number of tiles.
20383 This filter does the reverse of @ref{tile}.
20385 The filter accepts the following options:
20390 Set the grid size (i.e. the number of lines and columns). For the syntax of
20391 this option, check the
20392 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20395 @subsection Examples
20399 Produce a 1-second video from a still image file made of 25 frames stacked
20400 vertically, like an analogic film reel:
20402 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
20408 Apply ultra slow/simple postprocessing filter that compresses and decompresses
20409 the image at several (or - in the case of @option{quality} level @code{8} - all)
20410 shifts and average the results.
20412 The way this differs from the behavior of spp is that uspp actually encodes &
20413 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
20414 DCT similar to MJPEG.
20416 The filter accepts the following options:
20420 Set quality. This option defines the number of levels for averaging. It accepts
20421 an integer in the range 0-8. If set to @code{0}, the filter will have no
20422 effect. A value of @code{8} means the higher quality. For each increment of
20423 that value the speed drops by a factor of approximately 2. Default value is
20427 Force a constant quantization parameter. If not set, the filter will use the QP
20428 from the video stream (if available).
20433 Convert 360 videos between various formats.
20435 The filter accepts the following options:
20441 Set format of the input/output video.
20449 Equirectangular projection.
20454 Cubemap with 3x2/6x1/1x6 layout.
20456 Format specific options:
20461 Set padding proportion for the input/output cubemap. Values in decimals.
20468 1% of face is padding. For example, with 1920x1280 resolution face size would be 640x640 and padding would be 3 pixels from each side. (640 * 0.01 = 6 pixels)
20471 Default value is @b{@samp{0}}.
20472 Maximum value is @b{@samp{0.1}}.
20476 Set fixed padding for the input/output cubemap. Values in pixels.
20478 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
20482 Set order of faces for the input/output cubemap. Choose one direction for each position.
20484 Designation of directions:
20500 Default value is @b{@samp{rludfb}}.
20504 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
20506 Designation of angles:
20509 0 degrees clockwise
20511 90 degrees clockwise
20513 180 degrees clockwise
20515 270 degrees clockwise
20518 Default value is @b{@samp{000000}}.
20522 Equi-Angular Cubemap.
20529 Format specific options:
20534 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20536 If diagonal field of view is set it overrides horizontal and vertical field of view.
20541 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20543 If diagonal field of view is set it overrides horizontal and vertical field of view.
20549 Format specific options:
20554 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20556 If diagonal field of view is set it overrides horizontal and vertical field of view.
20561 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20563 If diagonal field of view is set it overrides horizontal and vertical field of view.
20569 Facebook's 360 formats.
20572 Stereographic format.
20574 Format specific options:
20579 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20581 If diagonal field of view is set it overrides horizontal and vertical field of view.
20586 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20588 If diagonal field of view is set it overrides horizontal and vertical field of view.
20595 Ball format, gives significant distortion toward the back.
20598 Hammer-Aitoff map projection format.
20601 Sinusoidal map projection format.
20604 Fisheye projection.
20606 Format specific options:
20611 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20613 If diagonal field of view is set it overrides horizontal and vertical field of view.
20618 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20620 If diagonal field of view is set it overrides horizontal and vertical field of view.
20624 Pannini projection.
20626 Format specific options:
20629 Set output pannini parameter.
20632 Set input pannini parameter.
20636 Cylindrical projection.
20638 Format specific options:
20643 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20645 If diagonal field of view is set it overrides horizontal and vertical field of view.
20650 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20652 If diagonal field of view is set it overrides horizontal and vertical field of view.
20656 Perspective projection. @i{(output only)}
20658 Format specific options:
20661 Set perspective parameter.
20665 Tetrahedron projection.
20668 Truncated square pyramid projection.
20672 Half equirectangular projection.
20677 Format specific options:
20682 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20684 If diagonal field of view is set it overrides horizontal and vertical field of view.
20689 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20691 If diagonal field of view is set it overrides horizontal and vertical field of view.
20695 Orthographic format.
20697 Format specific options:
20702 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20704 If diagonal field of view is set it overrides horizontal and vertical field of view.
20709 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20711 If diagonal field of view is set it overrides horizontal and vertical field of view.
20715 Octahedron projection.
20719 Set interpolation method.@*
20720 @i{Note: more complex interpolation methods require much more memory to run.}
20730 Bilinear interpolation.
20732 Lagrange9 interpolation.
20735 Bicubic interpolation.
20738 Lanczos interpolation.
20741 Spline16 interpolation.
20744 Gaussian interpolation.
20746 Mitchell interpolation.
20749 Default value is @b{@samp{line}}.
20753 Set the output video resolution.
20755 Default resolution depends on formats.
20759 Set the input/output stereo format.
20770 Default value is @b{@samp{2d}} for input and output format.
20775 Set rotation for the output video. Values in degrees.
20778 Set rotation order for the output video. Choose one item for each position.
20789 Default value is @b{@samp{ypr}}.
20794 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
20798 Set if input video is flipped horizontally/vertically. Boolean values.
20801 Set if input video is transposed. Boolean value, by default disabled.
20804 Set if output video needs to be transposed. Boolean value, by default disabled.
20807 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
20810 @subsection Examples
20814 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
20816 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
20819 Extract back view of Equi-Angular Cubemap:
20821 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
20824 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
20826 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
20830 @subsection Commands
20832 This filter supports subset of above options as @ref{commands}.
20834 @section vaguedenoiser
20836 Apply a wavelet based denoiser.
20838 It transforms each frame from the video input into the wavelet domain,
20839 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
20840 the obtained coefficients. It does an inverse wavelet transform after.
20841 Due to wavelet properties, it should give a nice smoothed result, and
20842 reduced noise, without blurring picture features.
20844 This filter accepts the following options:
20848 The filtering strength. The higher, the more filtered the video will be.
20849 Hard thresholding can use a higher threshold than soft thresholding
20850 before the video looks overfiltered. Default value is 2.
20853 The filtering method the filter will use.
20855 It accepts the following values:
20858 All values under the threshold will be zeroed.
20861 All values under the threshold will be zeroed. All values above will be
20862 reduced by the threshold.
20865 Scales or nullifies coefficients - intermediary between (more) soft and
20866 (less) hard thresholding.
20869 Default is garrote.
20872 Number of times, the wavelet will decompose the picture. Picture can't
20873 be decomposed beyond a particular point (typically, 8 for a 640x480
20874 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
20877 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
20880 A list of the planes to process. By default all planes are processed.
20883 The threshold type the filter will use.
20885 It accepts the following values:
20888 Threshold used is same for all decompositions.
20891 Threshold used depends also on each decomposition coefficients.
20894 Default is universal.
20897 @section vectorscope
20899 Display 2 color component values in the two dimensional graph (which is called
20902 This filter accepts the following options:
20906 Set vectorscope mode.
20908 It accepts the following values:
20912 Gray values are displayed on graph, higher brightness means more pixels have
20913 same component color value on location in graph. This is the default mode.
20916 Gray values are displayed on graph. Surrounding pixels values which are not
20917 present in video frame are drawn in gradient of 2 color components which are
20918 set by option @code{x} and @code{y}. The 3rd color component is static.
20921 Actual color components values present in video frame are displayed on graph.
20924 Similar as color2 but higher frequency of same values @code{x} and @code{y}
20925 on graph increases value of another color component, which is luminance by
20926 default values of @code{x} and @code{y}.
20929 Actual colors present in video frame are displayed on graph. If two different
20930 colors map to same position on graph then color with higher value of component
20931 not present in graph is picked.
20934 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
20935 component picked from radial gradient.
20939 Set which color component will be represented on X-axis. Default is @code{1}.
20942 Set which color component will be represented on Y-axis. Default is @code{2}.
20945 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
20946 of color component which represents frequency of (X, Y) location in graph.
20951 No envelope, this is default.
20954 Instant envelope, even darkest single pixel will be clearly highlighted.
20957 Hold maximum and minimum values presented in graph over time. This way you
20958 can still spot out of range values without constantly looking at vectorscope.
20961 Peak and instant envelope combined together.
20965 Set what kind of graticule to draw.
20974 Set graticule opacity.
20977 Set graticule flags.
20981 Draw graticule for white point.
20984 Draw graticule for black point.
20987 Draw color points short names.
20991 Set background opacity.
20993 @item lthreshold, l
20994 Set low threshold for color component not represented on X or Y axis.
20995 Values lower than this value will be ignored. Default is 0.
20996 Note this value is multiplied with actual max possible value one pixel component
20997 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
21000 @item hthreshold, h
21001 Set high threshold for color component not represented on X or Y axis.
21002 Values higher than this value will be ignored. Default is 1.
21003 Note this value is multiplied with actual max possible value one pixel component
21004 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
21005 is 0.9 * 255 = 230.
21007 @item colorspace, c
21008 Set what kind of colorspace to use when drawing graticule.
21018 Set color tint for gray/tint vectorscope mode. By default both options are zero.
21019 This means no tint, and output will remain gray.
21022 @anchor{vidstabdetect}
21023 @section vidstabdetect
21025 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
21026 @ref{vidstabtransform} for pass 2.
21028 This filter generates a file with relative translation and rotation
21029 transform information about subsequent frames, which is then used by
21030 the @ref{vidstabtransform} filter.
21032 To enable compilation of this filter you need to configure FFmpeg with
21033 @code{--enable-libvidstab}.
21035 This filter accepts the following options:
21039 Set the path to the file used to write the transforms information.
21040 Default value is @file{transforms.trf}.
21043 Set how shaky the video is and how quick the camera is. It accepts an
21044 integer in the range 1-10, a value of 1 means little shakiness, a
21045 value of 10 means strong shakiness. Default value is 5.
21048 Set the accuracy of the detection process. It must be a value in the
21049 range 1-15. A value of 1 means low accuracy, a value of 15 means high
21050 accuracy. Default value is 15.
21053 Set stepsize of the search process. The region around minimum is
21054 scanned with 1 pixel resolution. Default value is 6.
21057 Set minimum contrast. Below this value a local measurement field is
21058 discarded. Must be a floating point value in the range 0-1. Default
21062 Set reference frame number for tripod mode.
21064 If enabled, the motion of the frames is compared to a reference frame
21065 in the filtered stream, identified by the specified number. The idea
21066 is to compensate all movements in a more-or-less static scene and keep
21067 the camera view absolutely still.
21069 If set to 0, it is disabled. The frames are counted starting from 1.
21072 Show fields and transforms in the resulting frames. It accepts an
21073 integer in the range 0-2. Default value is 0, which disables any
21077 @subsection Examples
21081 Use default values:
21087 Analyze strongly shaky movie and put the results in file
21088 @file{mytransforms.trf}:
21090 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
21094 Visualize the result of internal transformations in the resulting
21097 vidstabdetect=show=1
21101 Analyze a video with medium shakiness using @command{ffmpeg}:
21103 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
21107 @anchor{vidstabtransform}
21108 @section vidstabtransform
21110 Video stabilization/deshaking: pass 2 of 2,
21111 see @ref{vidstabdetect} for pass 1.
21113 Read a file with transform information for each frame and
21114 apply/compensate them. Together with the @ref{vidstabdetect}
21115 filter this can be used to deshake videos. See also
21116 @url{http://public.hronopik.de/vid.stab}. It is important to also use
21117 the @ref{unsharp} filter, see below.
21119 To enable compilation of this filter you need to configure FFmpeg with
21120 @code{--enable-libvidstab}.
21122 @subsection Options
21126 Set path to the file used to read the transforms. Default value is
21127 @file{transforms.trf}.
21130 Set the number of frames (value*2 + 1) used for lowpass filtering the
21131 camera movements. Default value is 10.
21133 For example a number of 10 means that 21 frames are used (10 in the
21134 past and 10 in the future) to smoothen the motion in the video. A
21135 larger value leads to a smoother video, but limits the acceleration of
21136 the camera (pan/tilt movements). 0 is a special case where a static
21137 camera is simulated.
21140 Set the camera path optimization algorithm.
21142 Accepted values are:
21145 gaussian kernel low-pass filter on camera motion (default)
21147 averaging on transformations
21151 Set maximal number of pixels to translate frames. Default value is -1,
21155 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
21156 value is -1, meaning no limit.
21159 Specify how to deal with borders that may be visible due to movement
21162 Available values are:
21165 keep image information from previous frame (default)
21167 fill the border black
21171 Invert transforms if set to 1. Default value is 0.
21174 Consider transforms as relative to previous frame if set to 1,
21175 absolute if set to 0. Default value is 0.
21178 Set percentage to zoom. A positive value will result in a zoom-in
21179 effect, a negative value in a zoom-out effect. Default value is 0 (no
21183 Set optimal zooming to avoid borders.
21185 Accepted values are:
21190 optimal static zoom value is determined (only very strong movements
21191 will lead to visible borders) (default)
21193 optimal adaptive zoom value is determined (no borders will be
21194 visible), see @option{zoomspeed}
21197 Note that the value given at zoom is added to the one calculated here.
21200 Set percent to zoom maximally each frame (enabled when
21201 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
21205 Specify type of interpolation.
21207 Available values are:
21212 linear only horizontal
21214 linear in both directions (default)
21216 cubic in both directions (slow)
21220 Enable virtual tripod mode if set to 1, which is equivalent to
21221 @code{relative=0:smoothing=0}. Default value is 0.
21223 Use also @code{tripod} option of @ref{vidstabdetect}.
21226 Increase log verbosity if set to 1. Also the detected global motions
21227 are written to the temporary file @file{global_motions.trf}. Default
21231 @subsection Examples
21235 Use @command{ffmpeg} for a typical stabilization with default values:
21237 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
21240 Note the use of the @ref{unsharp} filter which is always recommended.
21243 Zoom in a bit more and load transform data from a given file:
21245 vidstabtransform=zoom=5:input="mytransforms.trf"
21249 Smoothen the video even more:
21251 vidstabtransform=smoothing=30
21257 Flip the input video vertically.
21259 For example, to vertically flip a video with @command{ffmpeg}:
21261 ffmpeg -i in.avi -vf "vflip" out.avi
21266 Detect variable frame rate video.
21268 This filter tries to detect if the input is variable or constant frame rate.
21270 At end it will output number of frames detected as having variable delta pts,
21271 and ones with constant delta pts.
21272 If there was frames with variable delta, than it will also show min, max and
21273 average delta encountered.
21277 Boost or alter saturation.
21279 The filter accepts the following options:
21282 Set strength of boost if positive value or strength of alter if negative value.
21283 Default is 0. Allowed range is from -2 to 2.
21286 Set the red balance. Default is 1. Allowed range is from -10 to 10.
21289 Set the green balance. Default is 1. Allowed range is from -10 to 10.
21292 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
21295 Set the red luma coefficient.
21298 Set the green luma coefficient.
21301 Set the blue luma coefficient.
21304 If @code{intensity} is negative and this is set to 1, colors will change,
21305 otherwise colors will be less saturated, more towards gray.
21308 @subsection Commands
21310 This filter supports the all above options as @ref{commands}.
21314 Obtain the average VIF (Visual Information Fidelity) between two input videos.
21316 This filter takes two input videos.
21318 Both input videos must have the same resolution and pixel format for
21319 this filter to work correctly. Also it assumes that both inputs
21320 have the same number of frames, which are compared one by one.
21322 The obtained average VIF score is printed through the logging system.
21324 The filter stores the calculated VIF score of each frame.
21326 In the below example the input file @file{main.mpg} being processed is compared
21327 with the reference file @file{ref.mpg}.
21330 ffmpeg -i main.mpg -i ref.mpg -lavfi vif -f null -
21336 Make or reverse a natural vignetting effect.
21338 The filter accepts the following options:
21342 Set lens angle expression as a number of radians.
21344 The value is clipped in the @code{[0,PI/2]} range.
21346 Default value: @code{"PI/5"}
21350 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
21354 Set forward/backward mode.
21356 Available modes are:
21359 The larger the distance from the central point, the darker the image becomes.
21362 The larger the distance from the central point, the brighter the image becomes.
21363 This can be used to reverse a vignette effect, though there is no automatic
21364 detection to extract the lens @option{angle} and other settings (yet). It can
21365 also be used to create a burning effect.
21368 Default value is @samp{forward}.
21371 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
21373 It accepts the following values:
21376 Evaluate expressions only once during the filter initialization.
21379 Evaluate expressions for each incoming frame. This is way slower than the
21380 @samp{init} mode since it requires all the scalers to be re-computed, but it
21381 allows advanced dynamic expressions.
21384 Default value is @samp{init}.
21387 Set dithering to reduce the circular banding effects. Default is @code{1}
21391 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
21392 Setting this value to the SAR of the input will make a rectangular vignetting
21393 following the dimensions of the video.
21395 Default is @code{1/1}.
21398 @subsection Expressions
21400 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
21401 following parameters.
21406 input width and height
21409 the number of input frame, starting from 0
21412 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
21413 @var{TB} units, NAN if undefined
21416 frame rate of the input video, NAN if the input frame rate is unknown
21419 the PTS (Presentation TimeStamp) of the filtered video frame,
21420 expressed in seconds, NAN if undefined
21423 time base of the input video
21427 @subsection Examples
21431 Apply simple strong vignetting effect:
21437 Make a flickering vignetting:
21439 vignette='PI/4+random(1)*PI/50':eval=frame
21444 @section vmafmotion
21446 Obtain the average VMAF motion score of a video.
21447 It is one of the component metrics of VMAF.
21449 The obtained average motion score is printed through the logging system.
21451 The filter accepts the following options:
21455 If specified, the filter will use the named file to save the motion score of
21456 each frame with respect to the previous frame.
21457 When filename equals "-" the data is sent to standard output.
21462 ffmpeg -i ref.mpg -vf vmafmotion -f null -
21466 Stack input videos vertically.
21468 All streams must be of same pixel format and of same width.
21470 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
21471 to create same output.
21473 The filter accepts the following options:
21477 Set number of input streams. Default is 2.
21480 If set to 1, force the output to terminate when the shortest input
21481 terminates. Default value is 0.
21486 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
21487 Deinterlacing Filter").
21489 Based on the process described by Martin Weston for BBC R&D, and
21490 implemented based on the de-interlace algorithm written by Jim
21491 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
21492 uses filter coefficients calculated by BBC R&D.
21494 This filter uses field-dominance information in frame to decide which
21495 of each pair of fields to place first in the output.
21496 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
21498 There are two sets of filter coefficients, so called "simple"
21499 and "complex". Which set of filter coefficients is used can
21500 be set by passing an optional parameter:
21504 Set the interlacing filter coefficients. Accepts one of the following values:
21508 Simple filter coefficient set.
21510 More-complex filter coefficient set.
21512 Default value is @samp{complex}.
21515 The interlacing mode to adopt. It accepts one of the following values:
21519 Output one frame for each frame.
21521 Output one frame for each field.
21524 The default value is @code{field}.
21527 The picture field parity assumed for the input interlaced video. It accepts one
21528 of the following values:
21532 Assume the top field is first.
21534 Assume the bottom field is first.
21536 Enable automatic detection of field parity.
21539 The default value is @code{auto}.
21540 If the interlacing is unknown or the decoder does not export this information,
21541 top field first will be assumed.
21544 Specify which frames to deinterlace. Accepts one of the following values:
21548 Deinterlace all frames,
21550 Only deinterlace frames marked as interlaced.
21553 Default value is @samp{all}.
21556 @subsection Commands
21557 This filter supports same @ref{commands} as options.
21560 Video waveform monitor.
21562 The waveform monitor plots color component intensity. By default luminance
21563 only. Each column of the waveform corresponds to a column of pixels in the
21566 It accepts the following options:
21570 Can be either @code{row}, or @code{column}. Default is @code{column}.
21571 In row mode, the graph on the left side represents color component value 0 and
21572 the right side represents value = 255. In column mode, the top side represents
21573 color component value = 0 and bottom side represents value = 255.
21576 Set intensity. Smaller values are useful to find out how many values of the same
21577 luminance are distributed across input rows/columns.
21578 Default value is @code{0.04}. Allowed range is [0, 1].
21581 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
21582 In mirrored mode, higher values will be represented on the left
21583 side for @code{row} mode and at the top for @code{column} mode. Default is
21584 @code{1} (mirrored).
21588 It accepts the following values:
21591 Presents information identical to that in the @code{parade}, except
21592 that the graphs representing color components are superimposed directly
21595 This display mode makes it easier to spot relative differences or similarities
21596 in overlapping areas of the color components that are supposed to be identical,
21597 such as neutral whites, grays, or blacks.
21600 Display separate graph for the color components side by side in
21601 @code{row} mode or one below the other in @code{column} mode.
21604 Display separate graph for the color components side by side in
21605 @code{column} mode or one below the other in @code{row} mode.
21607 Using this display mode makes it easy to spot color casts in the highlights
21608 and shadows of an image, by comparing the contours of the top and the bottom
21609 graphs of each waveform. Since whites, grays, and blacks are characterized
21610 by exactly equal amounts of red, green, and blue, neutral areas of the picture
21611 should display three waveforms of roughly equal width/height. If not, the
21612 correction is easy to perform by making level adjustments the three waveforms.
21614 Default is @code{stack}.
21616 @item components, c
21617 Set which color components to display. Default is 1, which means only luminance
21618 or red color component if input is in RGB colorspace. If is set for example to
21619 7 it will display all 3 (if) available color components.
21624 No envelope, this is default.
21627 Instant envelope, minimum and maximum values presented in graph will be easily
21628 visible even with small @code{step} value.
21631 Hold minimum and maximum values presented in graph across time. This way you
21632 can still spot out of range values without constantly looking at waveforms.
21635 Peak and instant envelope combined together.
21641 No filtering, this is default.
21644 Luma and chroma combined together.
21647 Similar as above, but shows difference between blue and red chroma.
21650 Similar as above, but use different colors.
21653 Similar as above, but again with different colors.
21656 Displays only chroma.
21659 Displays actual color value on waveform.
21662 Similar as above, but with luma showing frequency of chroma values.
21666 Set which graticule to display.
21670 Do not display graticule.
21673 Display green graticule showing legal broadcast ranges.
21676 Display orange graticule showing legal broadcast ranges.
21679 Display invert graticule showing legal broadcast ranges.
21683 Set graticule opacity.
21686 Set graticule flags.
21690 Draw numbers above lines. By default enabled.
21693 Draw dots instead of lines.
21697 Set scale used for displaying graticule.
21704 Default is digital.
21707 Set background opacity.
21711 Set tint for output.
21712 Only used with lowpass filter and when display is not overlay and input
21713 pixel formats are not RGB.
21716 @section weave, doubleweave
21718 The @code{weave} takes a field-based video input and join
21719 each two sequential fields into single frame, producing a new double
21720 height clip with half the frame rate and half the frame count.
21722 The @code{doubleweave} works same as @code{weave} but without
21723 halving frame rate and frame count.
21725 It accepts the following option:
21729 Set first field. Available values are:
21733 Set the frame as top-field-first.
21736 Set the frame as bottom-field-first.
21740 @subsection Examples
21744 Interlace video using @ref{select} and @ref{separatefields} filter:
21746 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
21751 Apply the xBR high-quality magnification filter which is designed for pixel
21752 art. It follows a set of edge-detection rules, see
21753 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
21755 It accepts the following option:
21759 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
21760 @code{3xBR} and @code{4} for @code{4xBR}.
21761 Default is @code{3}.
21766 Apply cross fade from one input video stream to another input video stream.
21767 The cross fade is applied for specified duration.
21769 The filter accepts the following options:
21773 Set one of available transition effects:
21821 Default transition effect is fade.
21824 Set cross fade duration in seconds.
21825 Default duration is 1 second.
21828 Set cross fade start relative to first input stream in seconds.
21829 Default offset is 0.
21832 Set expression for custom transition effect.
21834 The expressions can use the following variables and functions:
21839 The coordinates of the current sample.
21843 The width and height of the image.
21846 Progress of transition effect.
21849 Currently processed plane.
21852 Return value of first input at current location and plane.
21855 Return value of second input at current location and plane.
21861 Return the value of the pixel at location (@var{x},@var{y}) of the
21862 first/second/third/fourth component of first input.
21868 Return the value of the pixel at location (@var{x},@var{y}) of the
21869 first/second/third/fourth component of second input.
21873 @subsection Examples
21877 Cross fade from one input video to another input video, with fade transition and duration of transition
21878 of 2 seconds starting at offset of 5 seconds:
21880 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
21885 Pick median pixels from several input videos.
21887 The filter accepts the following options:
21891 Set number of inputs.
21892 Default is 3. Allowed range is from 3 to 255.
21893 If number of inputs is even number, than result will be mean value between two median values.
21896 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
21899 Set median percentile. Default value is @code{0.5}.
21900 Default value of @code{0.5} will pick always median values, while @code{0} will pick
21901 minimum values, and @code{1} maximum values.
21904 @subsection Commands
21906 This filter supports all above options as @ref{commands}, excluding option @code{inputs}.
21909 Stack video inputs into custom layout.
21911 All streams must be of same pixel format.
21913 The filter accepts the following options:
21917 Set number of input streams. Default is 2.
21920 Specify layout of inputs.
21921 This option requires the desired layout configuration to be explicitly set by the user.
21922 This sets position of each video input in output. Each input
21923 is separated by '|'.
21924 The first number represents the column, and the second number represents the row.
21925 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
21926 where X is video input from which to take width or height.
21927 Multiple values can be used when separated by '+'. In such
21928 case values are summed together.
21930 Note that if inputs are of different sizes gaps may appear, as not all of
21931 the output video frame will be filled. Similarly, videos can overlap each
21932 other if their position doesn't leave enough space for the full frame of
21935 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
21936 a layout must be set by the user.
21939 If set to 1, force the output to terminate when the shortest input
21940 terminates. Default value is 0.
21943 If set to valid color, all unused pixels will be filled with that color.
21944 By default fill is set to none, so it is disabled.
21947 @subsection Examples
21951 Display 4 inputs into 2x2 grid.
21955 input1(0, 0) | input3(w0, 0)
21956 input2(0, h0) | input4(w0, h0)
21960 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
21963 Note that if inputs are of different sizes, gaps or overlaps may occur.
21966 Display 4 inputs into 1x4 grid.
21973 input4(0, h0+h1+h2)
21977 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
21980 Note that if inputs are of different widths, unused space will appear.
21983 Display 9 inputs into 3x3 grid.
21987 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
21988 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
21989 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
21993 xstack=inputs=9:layout=0_0|0_h0|0_h0+h1|w0_0|w0_h0|w0_h0+h1|w0+w3_0|w0+w3_h0|w0+w3_h0+h1
21996 Note that if inputs are of different sizes, gaps or overlaps may occur.
21999 Display 16 inputs into 4x4 grid.
22003 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
22004 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
22005 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
22006 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
22010 xstack=inputs=16:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2|w0_0|w0_h0|w0_h0+h1|w0_h0+h1+h2|w0+w4_0|
22011 w0+w4_h0|w0+w4_h0+h1|w0+w4_h0+h1+h2|w0+w4+w8_0|w0+w4+w8_h0|w0+w4+w8_h0+h1|w0+w4+w8_h0+h1+h2
22014 Note that if inputs are of different sizes, gaps or overlaps may occur.
22021 Deinterlace the input video ("yadif" means "yet another deinterlacing
22024 It accepts the following parameters:
22030 The interlacing mode to adopt. It accepts one of the following values:
22033 @item 0, send_frame
22034 Output one frame for each frame.
22035 @item 1, send_field
22036 Output one frame for each field.
22037 @item 2, send_frame_nospatial
22038 Like @code{send_frame}, but it skips the spatial interlacing check.
22039 @item 3, send_field_nospatial
22040 Like @code{send_field}, but it skips the spatial interlacing check.
22043 The default value is @code{send_frame}.
22046 The picture field parity assumed for the input interlaced video. It accepts one
22047 of the following values:
22051 Assume the top field is first.
22053 Assume the bottom field is first.
22055 Enable automatic detection of field parity.
22058 The default value is @code{auto}.
22059 If the interlacing is unknown or the decoder does not export this information,
22060 top field first will be assumed.
22063 Specify which frames to deinterlace. Accepts one of the following
22068 Deinterlace all frames.
22069 @item 1, interlaced
22070 Only deinterlace frames marked as interlaced.
22073 The default value is @code{all}.
22076 @section yadif_cuda
22078 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
22079 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
22082 It accepts the following parameters:
22088 The interlacing mode to adopt. It accepts one of the following values:
22091 @item 0, send_frame
22092 Output one frame for each frame.
22093 @item 1, send_field
22094 Output one frame for each field.
22095 @item 2, send_frame_nospatial
22096 Like @code{send_frame}, but it skips the spatial interlacing check.
22097 @item 3, send_field_nospatial
22098 Like @code{send_field}, but it skips the spatial interlacing check.
22101 The default value is @code{send_frame}.
22104 The picture field parity assumed for the input interlaced video. It accepts one
22105 of the following values:
22109 Assume the top field is first.
22111 Assume the bottom field is first.
22113 Enable automatic detection of field parity.
22116 The default value is @code{auto}.
22117 If the interlacing is unknown or the decoder does not export this information,
22118 top field first will be assumed.
22121 Specify which frames to deinterlace. Accepts one of the following
22126 Deinterlace all frames.
22127 @item 1, interlaced
22128 Only deinterlace frames marked as interlaced.
22131 The default value is @code{all}.
22136 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
22137 The algorithm is described in
22138 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
22140 It accepts the following parameters:
22144 Set the window radius. Default value is 3.
22147 Set which planes to filter. Default is only the first plane.
22150 Set blur strength. Default value is 128.
22153 @subsection Commands
22154 This filter supports same @ref{commands} as options.
22158 Apply Zoom & Pan effect.
22160 This filter accepts the following options:
22164 Set the zoom expression. Range is 1-10. Default is 1.
22168 Set the x and y expression. Default is 0.
22171 Set the duration expression in number of frames.
22172 This sets for how many number of frames effect will last for
22173 single input image.
22176 Set the output image size, default is 'hd720'.
22179 Set the output frame rate, default is '25'.
22182 Each expression can contain the following constants:
22201 Output frame count.
22204 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
22206 @item out_time, time, ot
22207 The output timestamp expressed in seconds.
22211 Last calculated 'x' and 'y' position from 'x' and 'y' expression
22212 for current input frame.
22216 'x' and 'y' of last output frame of previous input frame or 0 when there was
22217 not yet such frame (first input frame).
22220 Last calculated zoom from 'z' expression for current input frame.
22223 Last calculated zoom of last output frame of previous input frame.
22226 Number of output frames for current input frame. Calculated from 'd' expression
22227 for each input frame.
22230 number of output frames created for previous input frame
22233 Rational number: input width / input height
22236 sample aspect ratio
22239 display aspect ratio
22243 @subsection Examples
22247 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
22249 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='if(gte(zoom,1.5),x,x+1/a)':y='if(gte(zoom,1.5),y,y+1)':s=640x360
22253 Zoom in up to 1.5x and pan always at center of picture:
22255 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22259 Same as above but without pausing:
22261 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22265 Zoom in 2x into center of picture only for the first second of the input video:
22267 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22274 Scale (resize) the input video, using the z.lib library:
22275 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
22276 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
22278 The zscale filter forces the output display aspect ratio to be the same
22279 as the input, by changing the output sample aspect ratio.
22281 If the input image format is different from the format requested by
22282 the next filter, the zscale filter will convert the input to the
22285 @subsection Options
22286 The filter accepts the following options.
22291 Set the output video dimension expression. Default value is the input
22294 If the @var{width} or @var{w} value is 0, the input width is used for
22295 the output. If the @var{height} or @var{h} value is 0, the input height
22296 is used for the output.
22298 If one and only one of the values is -n with n >= 1, the zscale filter
22299 will use a value that maintains the aspect ratio of the input image,
22300 calculated from the other specified dimension. After that it will,
22301 however, make sure that the calculated dimension is divisible by n and
22302 adjust the value if necessary.
22304 If both values are -n with n >= 1, the behavior will be identical to
22305 both values being set to 0 as previously detailed.
22307 See below for the list of accepted constants for use in the dimension
22311 Set the video size. For the syntax of this option, check the
22312 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22315 Set the dither type.
22317 Possible values are:
22322 @item error_diffusion
22328 Set the resize filter type.
22330 Possible values are:
22340 Default is bilinear.
22343 Set the color range.
22345 Possible values are:
22352 Default is same as input.
22355 Set the color primaries.
22357 Possible values are:
22367 Default is same as input.
22370 Set the transfer characteristics.
22372 Possible values are:
22386 Default is same as input.
22389 Set the colorspace matrix.
22391 Possible value are:
22402 Default is same as input.
22405 Set the input color range.
22407 Possible values are:
22414 Default is same as input.
22416 @item primariesin, pin
22417 Set the input color primaries.
22419 Possible values are:
22429 Default is same as input.
22431 @item transferin, tin
22432 Set the input transfer characteristics.
22434 Possible values are:
22445 Default is same as input.
22447 @item matrixin, min
22448 Set the input colorspace matrix.
22450 Possible value are:
22462 Set the output chroma location.
22464 Possible values are:
22475 @item chromalin, cin
22476 Set the input chroma location.
22478 Possible values are:
22490 Set the nominal peak luminance.
22493 Parameter A for scaling filters. Parameter "b" for bicubic, and the number of
22494 filter taps for lanczos.
22497 Parameter B for scaling filters. Parameter "c" for bicubic.
22500 The values of the @option{w} and @option{h} options are expressions
22501 containing the following constants:
22506 The input width and height
22510 These are the same as @var{in_w} and @var{in_h}.
22514 The output (scaled) width and height
22518 These are the same as @var{out_w} and @var{out_h}
22521 The same as @var{iw} / @var{ih}
22524 input sample aspect ratio
22527 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
22531 horizontal and vertical input chroma subsample values. For example for the
22532 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22536 horizontal and vertical output chroma subsample values. For example for the
22537 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22540 @subsection Commands
22542 This filter supports the following commands:
22546 Set the output video dimension expression.
22547 The command accepts the same syntax of the corresponding option.
22549 If the specified expression is not valid, it is kept at its current
22553 @c man end VIDEO FILTERS
22555 @chapter OpenCL Video Filters
22556 @c man begin OPENCL VIDEO FILTERS
22558 Below is a description of the currently available OpenCL video filters.
22560 To enable compilation of these filters you need to configure FFmpeg with
22561 @code{--enable-opencl}.
22563 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
22566 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
22567 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
22568 given device parameters.
22570 @item -filter_hw_device @var{name}
22571 Pass the hardware device called @var{name} to all filters in any filter graph.
22575 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
22579 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
22581 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
22585 Since OpenCL filters are not able to access frame data in normal memory, all frame data needs to be uploaded(@ref{hwupload}) to hardware surfaces connected to the appropriate device before being used and then downloaded(@ref{hwdownload}) back to normal memory. Note that @ref{hwupload} will upload to a surface with the same layout as the software frame, so it may be necessary to add a @ref{format} filter immediately before to get the input into the right format and @ref{hwdownload} does not support all formats on the output - it may be necessary to insert an additional @ref{format} filter immediately following in the graph to get the output in a supported format.
22587 @section avgblur_opencl
22589 Apply average blur filter.
22591 The filter accepts the following options:
22595 Set horizontal radius size.
22596 Range is @code{[1, 1024]} and default value is @code{1}.
22599 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22602 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
22605 @subsection Example
22609 Apply average blur filter with horizontal and vertical size of 3, setting each pixel of the output to the average value of the 7x7 region centered on it in the input. For pixels on the edges of the image, the region does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
22611 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
22615 @section boxblur_opencl
22617 Apply a boxblur algorithm to the input video.
22619 It accepts the following parameters:
22623 @item luma_radius, lr
22624 @item luma_power, lp
22625 @item chroma_radius, cr
22626 @item chroma_power, cp
22627 @item alpha_radius, ar
22628 @item alpha_power, ap
22632 A description of the accepted options follows.
22635 @item luma_radius, lr
22636 @item chroma_radius, cr
22637 @item alpha_radius, ar
22638 Set an expression for the box radius in pixels used for blurring the
22639 corresponding input plane.
22641 The radius value must be a non-negative number, and must not be
22642 greater than the value of the expression @code{min(w,h)/2} for the
22643 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
22646 Default value for @option{luma_radius} is "2". If not specified,
22647 @option{chroma_radius} and @option{alpha_radius} default to the
22648 corresponding value set for @option{luma_radius}.
22650 The expressions can contain the following constants:
22654 The input width and height in pixels.
22658 The input chroma image width and height in pixels.
22662 The horizontal and vertical chroma subsample values. For example, for the
22663 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
22666 @item luma_power, lp
22667 @item chroma_power, cp
22668 @item alpha_power, ap
22669 Specify how many times the boxblur filter is applied to the
22670 corresponding plane.
22672 Default value for @option{luma_power} is 2. If not specified,
22673 @option{chroma_power} and @option{alpha_power} default to the
22674 corresponding value set for @option{luma_power}.
22676 A value of 0 will disable the effect.
22679 @subsection Examples
22681 Apply boxblur filter, setting each pixel of the output to the average value of box-radiuses @var{luma_radius}, @var{chroma_radius}, @var{alpha_radius} for each plane respectively. The filter will apply @var{luma_power}, @var{chroma_power}, @var{alpha_power} times onto the corresponding plane. For pixels on the edges of the image, the radius does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
22685 Apply a boxblur filter with the luma, chroma, and alpha radius
22686 set to 2 and luma, chroma, and alpha power set to 3. The filter will run 3 times with box-radius set to 2 for every plane of the image.
22688 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
22689 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
22693 Apply a boxblur filter with luma radius set to 2, luma_power to 1, chroma_radius to 4, chroma_power to 5, alpha_radius to 3 and alpha_power to 7.
22695 For the luma plane, a 2x2 box radius will be run once.
22697 For the chroma plane, a 4x4 box radius will be run 5 times.
22699 For the alpha plane, a 3x3 box radius will be run 7 times.
22701 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
22705 @section colorkey_opencl
22706 RGB colorspace color keying.
22708 The filter accepts the following options:
22712 The color which will be replaced with transparency.
22715 Similarity percentage with the key color.
22717 0.01 matches only the exact key color, while 1.0 matches everything.
22722 0.0 makes pixels either fully transparent, or not transparent at all.
22724 Higher values result in semi-transparent pixels, with a higher transparency
22725 the more similar the pixels color is to the key color.
22728 @subsection Examples
22732 Make every semi-green pixel in the input transparent with some slight blending:
22734 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
22738 @section convolution_opencl
22740 Apply convolution of 3x3, 5x5, 7x7 matrix.
22742 The filter accepts the following options:
22749 Set matrix for each plane.
22750 Matrix is sequence of 9, 25 or 49 signed numbers.
22751 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
22757 Set multiplier for calculated value for each plane.
22758 If unset or 0, it will be sum of all matrix elements.
22759 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
22765 Set bias for each plane. This value is added to the result of the multiplication.
22766 Useful for making the overall image brighter or darker.
22767 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
22771 @subsection Examples
22777 -i INPUT -vf "hwupload, convolution_opencl=0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0, hwdownload" OUTPUT
22783 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9, hwdownload" OUTPUT
22787 Apply edge enhance:
22789 -i INPUT -vf "hwupload, convolution_opencl=0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128, hwdownload" OUTPUT
22795 -i INPUT -vf "hwupload, convolution_opencl=0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128, hwdownload" OUTPUT
22799 Apply laplacian edge detector which includes diagonals:
22801 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0, hwdownload" OUTPUT
22807 -i INPUT -vf "hwupload, convolution_opencl=-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2, hwdownload" OUTPUT
22811 @section erosion_opencl
22813 Apply erosion effect to the video.
22815 This filter replaces the pixel by the local(3x3) minimum.
22817 It accepts the following options:
22824 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22825 If @code{0}, plane will remain unchanged.
22828 Flag which specifies the pixel to refer to.
22829 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22831 Flags to local 3x3 coordinates region centered on @code{x}:
22840 @subsection Example
22844 Apply erosion filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local minimum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local minimum is more then threshold of the corresponding plane, output pixel will be set to input pixel - threshold of corresponding plane.
22846 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22850 @section deshake_opencl
22851 Feature-point based video stabilization filter.
22853 The filter accepts the following options:
22857 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
22860 Whether or not additional debug info should be displayed, both in the processed output and in the console.
22862 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
22864 Viewing point matches in the output video is only supported for RGB input.
22866 Defaults to @code{0}.
22868 @item adaptive_crop
22869 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
22871 Defaults to @code{1}.
22873 @item refine_features
22874 Whether or not feature points should be refined at a sub-pixel level.
22876 This can be turned off for a slight performance gain at the cost of precision.
22878 Defaults to @code{1}.
22880 @item smooth_strength
22881 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
22883 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
22885 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
22887 Defaults to @code{0.0}.
22889 @item smooth_window_multiplier
22890 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
22892 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
22894 Acceptable values range from @code{0.1} to @code{10.0}.
22896 Larger values increase the amount of motion data available for determining how to smooth the camera path,
22897 potentially improving smoothness, but also increase latency and memory usage.
22899 Defaults to @code{2.0}.
22903 @subsection Examples
22907 Stabilize a video with a fixed, medium smoothing strength:
22909 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
22913 Stabilize a video with debugging (both in console and in rendered video):
22915 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
22919 @section dilation_opencl
22921 Apply dilation effect to the video.
22923 This filter replaces the pixel by the local(3x3) maximum.
22925 It accepts the following options:
22932 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22933 If @code{0}, plane will remain unchanged.
22936 Flag which specifies the pixel to refer to.
22937 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22939 Flags to local 3x3 coordinates region centered on @code{x}:
22948 @subsection Example
22952 Apply dilation filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local maximum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local maximum is more then threshold of the corresponding plane, output pixel will be set to input pixel + threshold of corresponding plane.
22954 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22958 @section nlmeans_opencl
22960 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
22962 @section overlay_opencl
22964 Overlay one video on top of another.
22966 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
22967 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
22969 The filter accepts the following options:
22974 Set the x coordinate of the overlaid video on the main video.
22975 Default value is @code{0}.
22978 Set the y coordinate of the overlaid video on the main video.
22979 Default value is @code{0}.
22983 @subsection Examples
22987 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
22989 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22992 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
22994 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22999 @section pad_opencl
23001 Add paddings to the input image, and place the original input at the
23002 provided @var{x}, @var{y} coordinates.
23004 It accepts the following options:
23009 Specify an expression for the size of the output image with the
23010 paddings added. If the value for @var{width} or @var{height} is 0, the
23011 corresponding input size is used for the output.
23013 The @var{width} expression can reference the value set by the
23014 @var{height} expression, and vice versa.
23016 The default value of @var{width} and @var{height} is 0.
23020 Specify the offsets to place the input image at within the padded area,
23021 with respect to the top/left border of the output image.
23023 The @var{x} expression can reference the value set by the @var{y}
23024 expression, and vice versa.
23026 The default value of @var{x} and @var{y} is 0.
23028 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
23029 so the input image is centered on the padded area.
23032 Specify the color of the padded area. For the syntax of this option,
23033 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
23034 manual,ffmpeg-utils}.
23037 Pad to an aspect instead to a resolution.
23040 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
23041 options are expressions containing the following constants:
23046 The input video width and height.
23050 These are the same as @var{in_w} and @var{in_h}.
23054 The output width and height (the size of the padded area), as
23055 specified by the @var{width} and @var{height} expressions.
23059 These are the same as @var{out_w} and @var{out_h}.
23063 The x and y offsets as specified by the @var{x} and @var{y}
23064 expressions, or NAN if not yet specified.
23067 same as @var{iw} / @var{ih}
23070 input sample aspect ratio
23073 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
23076 @section prewitt_opencl
23078 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
23080 The filter accepts the following option:
23084 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23087 Set value which will be multiplied with filtered result.
23088 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23091 Set value which will be added to filtered result.
23092 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23095 @subsection Example
23099 Apply the Prewitt operator with scale set to 2 and delta set to 10.
23101 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
23105 @anchor{program_opencl}
23106 @section program_opencl
23108 Filter video using an OpenCL program.
23113 OpenCL program source file.
23116 Kernel name in program.
23119 Number of inputs to the filter. Defaults to 1.
23122 Size of output frames. Defaults to the same as the first input.
23126 The @code{program_opencl} filter also supports the @ref{framesync} options.
23128 The program source file must contain a kernel function with the given name,
23129 which will be run once for each plane of the output. Each run on a plane
23130 gets enqueued as a separate 2D global NDRange with one work-item for each
23131 pixel to be generated. The global ID offset for each work-item is therefore
23132 the coordinates of a pixel in the destination image.
23134 The kernel function needs to take the following arguments:
23137 Destination image, @var{__write_only image2d_t}.
23139 This image will become the output; the kernel should write all of it.
23141 Frame index, @var{unsigned int}.
23143 This is a counter starting from zero and increasing by one for each frame.
23145 Source images, @var{__read_only image2d_t}.
23147 These are the most recent images on each input. The kernel may read from
23148 them to generate the output, but they can't be written to.
23155 Copy the input to the output (output must be the same size as the input).
23157 __kernel void copy(__write_only image2d_t destination,
23158 unsigned int index,
23159 __read_only image2d_t source)
23161 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
23163 int2 location = (int2)(get_global_id(0), get_global_id(1));
23165 float4 value = read_imagef(source, sampler, location);
23167 write_imagef(destination, location, value);
23172 Apply a simple transformation, rotating the input by an amount increasing
23173 with the index counter. Pixel values are linearly interpolated by the
23174 sampler, and the output need not have the same dimensions as the input.
23176 __kernel void rotate_image(__write_only image2d_t dst,
23177 unsigned int index,
23178 __read_only image2d_t src)
23180 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23181 CLK_FILTER_LINEAR);
23183 float angle = (float)index / 100.0f;
23185 float2 dst_dim = convert_float2(get_image_dim(dst));
23186 float2 src_dim = convert_float2(get_image_dim(src));
23188 float2 dst_cen = dst_dim / 2.0f;
23189 float2 src_cen = src_dim / 2.0f;
23191 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23193 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
23195 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
23196 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
23198 src_pos = src_pos * src_dim / dst_dim;
23200 float2 src_loc = src_pos + src_cen;
23202 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
23203 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
23204 write_imagef(dst, dst_loc, 0.5f);
23206 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
23211 Blend two inputs together, with the amount of each input used varying
23212 with the index counter.
23214 __kernel void blend_images(__write_only image2d_t dst,
23215 unsigned int index,
23216 __read_only image2d_t src1,
23217 __read_only image2d_t src2)
23219 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23220 CLK_FILTER_LINEAR);
23222 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
23224 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23225 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
23226 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
23228 float4 val1 = read_imagef(src1, sampler, src1_loc);
23229 float4 val2 = read_imagef(src2, sampler, src2_loc);
23231 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
23237 @section roberts_opencl
23238 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
23240 The filter accepts the following option:
23244 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23247 Set value which will be multiplied with filtered result.
23248 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23251 Set value which will be added to filtered result.
23252 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23255 @subsection Example
23259 Apply the Roberts cross operator with scale set to 2 and delta set to 10
23261 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
23265 @section sobel_opencl
23267 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
23269 The filter accepts the following option:
23273 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23276 Set value which will be multiplied with filtered result.
23277 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23280 Set value which will be added to filtered result.
23281 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23284 @subsection Example
23288 Apply sobel operator with scale set to 2 and delta set to 10
23290 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
23294 @section tonemap_opencl
23296 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
23298 It accepts the following parameters:
23302 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
23305 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
23308 Apply desaturation for highlights that exceed this level of brightness. The
23309 higher the parameter, the more color information will be preserved. This
23310 setting helps prevent unnaturally blown-out colors for super-highlights, by
23311 (smoothly) turning into white instead. This makes images feel more natural,
23312 at the cost of reducing information about out-of-range colors.
23314 The default value is 0.5, and the algorithm here is a little different from
23315 the cpu version tonemap currently. A setting of 0.0 disables this option.
23318 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
23319 is used to detect whether the scene has changed or not. If the distance between
23320 the current frame average brightness and the current running average exceeds
23321 a threshold value, we would re-calculate scene average and peak brightness.
23322 The default value is 0.2.
23325 Specify the output pixel format.
23327 Currently supported formats are:
23334 Set the output color range.
23336 Possible values are:
23342 Default is same as input.
23345 Set the output color primaries.
23347 Possible values are:
23353 Default is same as input.
23356 Set the output transfer characteristics.
23358 Possible values are:
23367 Set the output colorspace matrix.
23369 Possible value are:
23375 Default is same as input.
23379 @subsection Example
23383 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
23385 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
23389 @section unsharp_opencl
23391 Sharpen or blur the input video.
23393 It accepts the following parameters:
23396 @item luma_msize_x, lx
23397 Set the luma matrix horizontal size.
23398 Range is @code{[1, 23]} and default value is @code{5}.
23400 @item luma_msize_y, ly
23401 Set the luma matrix vertical size.
23402 Range is @code{[1, 23]} and default value is @code{5}.
23404 @item luma_amount, la
23405 Set the luma effect strength.
23406 Range is @code{[-10, 10]} and default value is @code{1.0}.
23408 Negative values will blur the input video, while positive values will
23409 sharpen it, a value of zero will disable the effect.
23411 @item chroma_msize_x, cx
23412 Set the chroma matrix horizontal size.
23413 Range is @code{[1, 23]} and default value is @code{5}.
23415 @item chroma_msize_y, cy
23416 Set the chroma matrix vertical size.
23417 Range is @code{[1, 23]} and default value is @code{5}.
23419 @item chroma_amount, ca
23420 Set the chroma effect strength.
23421 Range is @code{[-10, 10]} and default value is @code{0.0}.
23423 Negative values will blur the input video, while positive values will
23424 sharpen it, a value of zero will disable the effect.
23428 All parameters are optional and default to the equivalent of the
23429 string '5:5:1.0:5:5:0.0'.
23431 @subsection Examples
23435 Apply strong luma sharpen effect:
23437 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
23441 Apply a strong blur of both luma and chroma parameters:
23443 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
23447 @section xfade_opencl
23449 Cross fade two videos with custom transition effect by using OpenCL.
23451 It accepts the following options:
23455 Set one of possible transition effects.
23459 Select custom transition effect, the actual transition description
23460 will be picked from source and kernel options.
23472 Default transition is fade.
23476 OpenCL program source file for custom transition.
23479 Set name of kernel to use for custom transition from program source file.
23482 Set duration of video transition.
23485 Set time of start of transition relative to first video.
23488 The program source file must contain a kernel function with the given name,
23489 which will be run once for each plane of the output. Each run on a plane
23490 gets enqueued as a separate 2D global NDRange with one work-item for each
23491 pixel to be generated. The global ID offset for each work-item is therefore
23492 the coordinates of a pixel in the destination image.
23494 The kernel function needs to take the following arguments:
23497 Destination image, @var{__write_only image2d_t}.
23499 This image will become the output; the kernel should write all of it.
23502 First Source image, @var{__read_only image2d_t}.
23503 Second Source image, @var{__read_only image2d_t}.
23505 These are the most recent images on each input. The kernel may read from
23506 them to generate the output, but they can't be written to.
23509 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
23516 Apply dots curtain transition effect:
23518 __kernel void blend_images(__write_only image2d_t dst,
23519 __read_only image2d_t src1,
23520 __read_only image2d_t src2,
23523 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23524 CLK_FILTER_LINEAR);
23525 int2 p = (int2)(get_global_id(0), get_global_id(1));
23526 float2 rp = (float2)(get_global_id(0), get_global_id(1));
23527 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
23530 float2 dots = (float2)(20.0, 20.0);
23531 float2 center = (float2)(0,0);
23534 float4 val1 = read_imagef(src1, sampler, p);
23535 float4 val2 = read_imagef(src2, sampler, p);
23536 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
23538 write_imagef(dst, p, next ? val1 : val2);
23544 @c man end OPENCL VIDEO FILTERS
23546 @chapter VAAPI Video Filters
23547 @c man begin VAAPI VIDEO FILTERS
23549 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
23551 To enable compilation of these filters you need to configure FFmpeg with
23552 @code{--enable-vaapi}.
23554 To use vaapi filters, you need to setup the vaapi device correctly. For more information, please read @url{https://trac.ffmpeg.org/wiki/Hardware/VAAPI}
23556 @section tonemap_vaapi
23558 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
23559 It maps the dynamic range of HDR10 content to the SDR content.
23560 It currently only accepts HDR10 as input.
23562 It accepts the following parameters:
23566 Specify the output pixel format.
23568 Currently supported formats are:
23577 Set the output color primaries.
23579 Default is same as input.
23582 Set the output transfer characteristics.
23587 Set the output colorspace matrix.
23589 Default is same as input.
23593 @subsection Example
23597 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
23599 tonemap_vaapi=format=p010:t=bt2020-10
23603 @c man end VAAPI VIDEO FILTERS
23605 @chapter Video Sources
23606 @c man begin VIDEO SOURCES
23608 Below is a description of the currently available video sources.
23612 Buffer video frames, and make them available to the filter chain.
23614 This source is mainly intended for a programmatic use, in particular
23615 through the interface defined in @file{libavfilter/buffersrc.h}.
23617 It accepts the following parameters:
23622 Specify the size (width and height) of the buffered video frames. For the
23623 syntax of this option, check the
23624 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23627 The input video width.
23630 The input video height.
23633 A string representing the pixel format of the buffered video frames.
23634 It may be a number corresponding to a pixel format, or a pixel format
23638 Specify the timebase assumed by the timestamps of the buffered frames.
23641 Specify the frame rate expected for the video stream.
23643 @item pixel_aspect, sar
23644 The sample (pixel) aspect ratio of the input video.
23647 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
23648 to the filtergraph description to specify swscale flags for automatically
23649 inserted scalers. See @ref{Filtergraph syntax}.
23651 @item hw_frames_ctx
23652 When using a hardware pixel format, this should be a reference to an
23653 AVHWFramesContext describing input frames.
23658 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
23661 will instruct the source to accept video frames with size 320x240 and
23662 with format "yuv410p", assuming 1/24 as the timestamps timebase and
23663 square pixels (1:1 sample aspect ratio).
23664 Since the pixel format with name "yuv410p" corresponds to the number 6
23665 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
23666 this example corresponds to:
23668 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
23671 Alternatively, the options can be specified as a flat string, but this
23672 syntax is deprecated:
23674 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
23678 Create a pattern generated by an elementary cellular automaton.
23680 The initial state of the cellular automaton can be defined through the
23681 @option{filename} and @option{pattern} options. If such options are
23682 not specified an initial state is created randomly.
23684 At each new frame a new row in the video is filled with the result of
23685 the cellular automaton next generation. The behavior when the whole
23686 frame is filled is defined by the @option{scroll} option.
23688 This source accepts the following options:
23692 Read the initial cellular automaton state, i.e. the starting row, from
23693 the specified file.
23694 In the file, each non-whitespace character is considered an alive
23695 cell, a newline will terminate the row, and further characters in the
23696 file will be ignored.
23699 Read the initial cellular automaton state, i.e. the starting row, from
23700 the specified string.
23702 Each non-whitespace character in the string is considered an alive
23703 cell, a newline will terminate the row, and further characters in the
23704 string will be ignored.
23707 Set the video rate, that is the number of frames generated per second.
23710 @item random_fill_ratio, ratio
23711 Set the random fill ratio for the initial cellular automaton row. It
23712 is a floating point number value ranging from 0 to 1, defaults to
23715 This option is ignored when a file or a pattern is specified.
23717 @item random_seed, seed
23718 Set the seed for filling randomly the initial row, must be an integer
23719 included between 0 and UINT32_MAX. If not specified, or if explicitly
23720 set to -1, the filter will try to use a good random seed on a best
23724 Set the cellular automaton rule, it is a number ranging from 0 to 255.
23725 Default value is 110.
23728 Set the size of the output video. For the syntax of this option, check the
23729 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23731 If @option{filename} or @option{pattern} is specified, the size is set
23732 by default to the width of the specified initial state row, and the
23733 height is set to @var{width} * PHI.
23735 If @option{size} is set, it must contain the width of the specified
23736 pattern string, and the specified pattern will be centered in the
23739 If a filename or a pattern string is not specified, the size value
23740 defaults to "320x518" (used for a randomly generated initial state).
23743 If set to 1, scroll the output upward when all the rows in the output
23744 have been already filled. If set to 0, the new generated row will be
23745 written over the top row just after the bottom row is filled.
23748 @item start_full, full
23749 If set to 1, completely fill the output with generated rows before
23750 outputting the first frame.
23751 This is the default behavior, for disabling set the value to 0.
23754 If set to 1, stitch the left and right row edges together.
23755 This is the default behavior, for disabling set the value to 0.
23758 @subsection Examples
23762 Read the initial state from @file{pattern}, and specify an output of
23765 cellauto=f=pattern:s=200x400
23769 Generate a random initial row with a width of 200 cells, with a fill
23772 cellauto=ratio=2/3:s=200x200
23776 Create a pattern generated by rule 18 starting by a single alive cell
23777 centered on an initial row with width 100:
23779 cellauto=p=@@:s=100x400:full=0:rule=18
23783 Specify a more elaborated initial pattern:
23785 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
23790 @anchor{coreimagesrc}
23791 @section coreimagesrc
23792 Video source generated on GPU using Apple's CoreImage API on OSX.
23794 This video source is a specialized version of the @ref{coreimage} video filter.
23795 Use a core image generator at the beginning of the applied filterchain to
23796 generate the content.
23798 The coreimagesrc video source accepts the following options:
23800 @item list_generators
23801 List all available generators along with all their respective options as well as
23802 possible minimum and maximum values along with the default values.
23804 list_generators=true
23808 Specify the size of the sourced video. For the syntax of this option, check the
23809 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23810 The default value is @code{320x240}.
23813 Specify the frame rate of the sourced video, as the number of frames
23814 generated per second. It has to be a string in the format
23815 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23816 number or a valid video frame rate abbreviation. The default value is
23820 Set the sample aspect ratio of the sourced video.
23823 Set the duration of the sourced video. See
23824 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23825 for the accepted syntax.
23827 If not specified, or the expressed duration is negative, the video is
23828 supposed to be generated forever.
23831 Additionally, all options of the @ref{coreimage} video filter are accepted.
23832 A complete filterchain can be used for further processing of the
23833 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
23834 and examples for details.
23836 @subsection Examples
23841 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
23842 given as complete and escaped command-line for Apple's standard bash shell:
23844 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
23846 This example is equivalent to the QRCode example of @ref{coreimage} without the
23847 need for a nullsrc video source.
23852 Generate several gradients.
23856 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23857 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23860 Set frame rate, expressed as number of frames per second. Default
23863 @item c0, c1, c2, c3, c4, c5, c6, c7
23864 Set 8 colors. Default values for colors is to pick random one.
23866 @item x0, y0, y0, y1
23867 Set gradient line source and destination points. If negative or out of range, random ones
23871 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
23874 Set seed for picking gradient line points.
23877 Set the duration of the sourced video. See
23878 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23879 for the accepted syntax.
23881 If not specified, or the expressed duration is negative, the video is
23882 supposed to be generated forever.
23885 Set speed of gradients rotation.
23889 @section mandelbrot
23891 Generate a Mandelbrot set fractal, and progressively zoom towards the
23892 point specified with @var{start_x} and @var{start_y}.
23894 This source accepts the following options:
23899 Set the terminal pts value. Default value is 400.
23902 Set the terminal scale value.
23903 Must be a floating point value. Default value is 0.3.
23906 Set the inner coloring mode, that is the algorithm used to draw the
23907 Mandelbrot fractal internal region.
23909 It shall assume one of the following values:
23914 Show time until convergence.
23916 Set color based on point closest to the origin of the iterations.
23921 Default value is @var{mincol}.
23924 Set the bailout value. Default value is 10.0.
23927 Set the maximum of iterations performed by the rendering
23928 algorithm. Default value is 7189.
23931 Set outer coloring mode.
23932 It shall assume one of following values:
23934 @item iteration_count
23935 Set iteration count mode.
23936 @item normalized_iteration_count
23937 set normalized iteration count mode.
23939 Default value is @var{normalized_iteration_count}.
23942 Set frame rate, expressed as number of frames per second. Default
23946 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23947 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23950 Set the initial scale value. Default value is 3.0.
23953 Set the initial x position. Must be a floating point value between
23954 -100 and 100. Default value is -0.743643887037158704752191506114774.
23957 Set the initial y position. Must be a floating point value between
23958 -100 and 100. Default value is -0.131825904205311970493132056385139.
23963 Generate various test patterns, as generated by the MPlayer test filter.
23965 The size of the generated video is fixed, and is 256x256.
23966 This source is useful in particular for testing encoding features.
23968 This source accepts the following options:
23973 Specify the frame rate of the sourced video, as the number of frames
23974 generated per second. It has to be a string in the format
23975 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23976 number or a valid video frame rate abbreviation. The default value is
23980 Set the duration of the sourced video. See
23981 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23982 for the accepted syntax.
23984 If not specified, or the expressed duration is negative, the video is
23985 supposed to be generated forever.
23989 Set the number or the name of the test to perform. Supported tests are:
24003 @item max_frames, m
24004 Set the maximum number of frames generated for each test, default value is 30.
24008 Default value is "all", which will cycle through the list of all tests.
24013 mptestsrc=t=dc_luma
24016 will generate a "dc_luma" test pattern.
24018 @section frei0r_src
24020 Provide a frei0r source.
24022 To enable compilation of this filter you need to install the frei0r
24023 header and configure FFmpeg with @code{--enable-frei0r}.
24025 This source accepts the following parameters:
24030 The size of the video to generate. For the syntax of this option, check the
24031 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24034 The framerate of the generated video. It may be a string of the form
24035 @var{num}/@var{den} or a frame rate abbreviation.
24038 The name to the frei0r source to load. For more information regarding frei0r and
24039 how to set the parameters, read the @ref{frei0r} section in the video filters
24042 @item filter_params
24043 A '|'-separated list of parameters to pass to the frei0r source.
24047 For example, to generate a frei0r partik0l source with size 200x200
24048 and frame rate 10 which is overlaid on the overlay filter main input:
24050 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
24055 Generate a life pattern.
24057 This source is based on a generalization of John Conway's life game.
24059 The sourced input represents a life grid, each pixel represents a cell
24060 which can be in one of two possible states, alive or dead. Every cell
24061 interacts with its eight neighbours, which are the cells that are
24062 horizontally, vertically, or diagonally adjacent.
24064 At each interaction the grid evolves according to the adopted rule,
24065 which specifies the number of neighbor alive cells which will make a
24066 cell stay alive or born. The @option{rule} option allows one to specify
24069 This source accepts the following options:
24073 Set the file from which to read the initial grid state. In the file,
24074 each non-whitespace character is considered an alive cell, and newline
24075 is used to delimit the end of each row.
24077 If this option is not specified, the initial grid is generated
24081 Set the video rate, that is the number of frames generated per second.
24084 @item random_fill_ratio, ratio
24085 Set the random fill ratio for the initial random grid. It is a
24086 floating point number value ranging from 0 to 1, defaults to 1/PHI.
24087 It is ignored when a file is specified.
24089 @item random_seed, seed
24090 Set the seed for filling the initial random grid, must be an integer
24091 included between 0 and UINT32_MAX. If not specified, or if explicitly
24092 set to -1, the filter will try to use a good random seed on a best
24098 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
24099 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
24100 @var{NS} specifies the number of alive neighbor cells which make a
24101 live cell stay alive, and @var{NB} the number of alive neighbor cells
24102 which make a dead cell to become alive (i.e. to "born").
24103 "s" and "b" can be used in place of "S" and "B", respectively.
24105 Alternatively a rule can be specified by an 18-bits integer. The 9
24106 high order bits are used to encode the next cell state if it is alive
24107 for each number of neighbor alive cells, the low order bits specify
24108 the rule for "borning" new cells. Higher order bits encode for an
24109 higher number of neighbor cells.
24110 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
24111 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
24113 Default value is "S23/B3", which is the original Conway's game of life
24114 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
24115 cells, and will born a new cell if there are three alive cells around
24119 Set the size of the output video. For the syntax of this option, check the
24120 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24122 If @option{filename} is specified, the size is set by default to the
24123 same size of the input file. If @option{size} is set, it must contain
24124 the size specified in the input file, and the initial grid defined in
24125 that file is centered in the larger resulting area.
24127 If a filename is not specified, the size value defaults to "320x240"
24128 (used for a randomly generated initial grid).
24131 If set to 1, stitch the left and right grid edges together, and the
24132 top and bottom edges also. Defaults to 1.
24135 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
24136 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
24137 value from 0 to 255.
24140 Set the color of living (or new born) cells.
24143 Set the color of dead cells. If @option{mold} is set, this is the first color
24144 used to represent a dead cell.
24147 Set mold color, for definitely dead and moldy cells.
24149 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
24150 ffmpeg-utils manual,ffmpeg-utils}.
24153 @subsection Examples
24157 Read a grid from @file{pattern}, and center it on a grid of size
24160 life=f=pattern:s=300x300
24164 Generate a random grid of size 200x200, with a fill ratio of 2/3:
24166 life=ratio=2/3:s=200x200
24170 Specify a custom rule for evolving a randomly generated grid:
24176 Full example with slow death effect (mold) using @command{ffplay}:
24178 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
24185 @anchor{haldclutsrc}
24188 @anchor{pal100bars}
24189 @anchor{rgbtestsrc}
24191 @anchor{smptehdbars}
24194 @anchor{yuvtestsrc}
24195 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
24197 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
24199 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
24201 The @code{color} source provides an uniformly colored input.
24203 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
24204 @ref{haldclut} filter.
24206 The @code{nullsrc} source returns unprocessed video frames. It is
24207 mainly useful to be employed in analysis / debugging tools, or as the
24208 source for filters which ignore the input data.
24210 The @code{pal75bars} source generates a color bars pattern, based on
24211 EBU PAL recommendations with 75% color levels.
24213 The @code{pal100bars} source generates a color bars pattern, based on
24214 EBU PAL recommendations with 100% color levels.
24216 The @code{rgbtestsrc} source generates an RGB test pattern useful for
24217 detecting RGB vs BGR issues. You should see a red, green and blue
24218 stripe from top to bottom.
24220 The @code{smptebars} source generates a color bars pattern, based on
24221 the SMPTE Engineering Guideline EG 1-1990.
24223 The @code{smptehdbars} source generates a color bars pattern, based on
24224 the SMPTE RP 219-2002.
24226 The @code{testsrc} source generates a test video pattern, showing a
24227 color pattern, a scrolling gradient and a timestamp. This is mainly
24228 intended for testing purposes.
24230 The @code{testsrc2} source is similar to testsrc, but supports more
24231 pixel formats instead of just @code{rgb24}. This allows using it as an
24232 input for other tests without requiring a format conversion.
24234 The @code{yuvtestsrc} source generates an YUV test pattern. You should
24235 see a y, cb and cr stripe from top to bottom.
24237 The sources accept the following parameters:
24242 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
24243 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
24244 pixels to be used as identity matrix for 3D lookup tables. Each component is
24245 coded on a @code{1/(N*N)} scale.
24248 Specify the color of the source, only available in the @code{color}
24249 source. For the syntax of this option, check the
24250 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
24253 Specify the size of the sourced video. For the syntax of this option, check the
24254 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24255 The default value is @code{320x240}.
24257 This option is not available with the @code{allrgb}, @code{allyuv}, and
24258 @code{haldclutsrc} filters.
24261 Specify the frame rate of the sourced video, as the number of frames
24262 generated per second. It has to be a string in the format
24263 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
24264 number or a valid video frame rate abbreviation. The default value is
24268 Set the duration of the sourced video. See
24269 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
24270 for the accepted syntax.
24272 If not specified, or the expressed duration is negative, the video is
24273 supposed to be generated forever.
24275 Since the frame rate is used as time base, all frames including the last one
24276 will have their full duration. If the specified duration is not a multiple
24277 of the frame duration, it will be rounded up.
24280 Set the sample aspect ratio of the sourced video.
24283 Specify the alpha (opacity) of the background, only available in the
24284 @code{testsrc2} source. The value must be between 0 (fully transparent) and
24285 255 (fully opaque, the default).
24288 Set the number of decimals to show in the timestamp, only available in the
24289 @code{testsrc} source.
24291 The displayed timestamp value will correspond to the original
24292 timestamp value multiplied by the power of 10 of the specified
24293 value. Default value is 0.
24296 @subsection Examples
24300 Generate a video with a duration of 5.3 seconds, with size
24301 176x144 and a frame rate of 10 frames per second:
24303 testsrc=duration=5.3:size=qcif:rate=10
24307 The following graph description will generate a red source
24308 with an opacity of 0.2, with size "qcif" and a frame rate of 10
24311 color=c=red@@0.2:s=qcif:r=10
24315 If the input content is to be ignored, @code{nullsrc} can be used. The
24316 following command generates noise in the luminance plane by employing
24317 the @code{geq} filter:
24319 nullsrc=s=256x256, geq=random(1)*255:128:128
24323 @subsection Commands
24325 The @code{color} source supports the following commands:
24329 Set the color of the created image. Accepts the same syntax of the
24330 corresponding @option{color} option.
24335 Generate video using an OpenCL program.
24340 OpenCL program source file.
24343 Kernel name in program.
24346 Size of frames to generate. This must be set.
24349 Pixel format to use for the generated frames. This must be set.
24352 Number of frames generated every second. Default value is '25'.
24356 For details of how the program loading works, see the @ref{program_opencl}
24363 Generate a colour ramp by setting pixel values from the position of the pixel
24364 in the output image. (Note that this will work with all pixel formats, but
24365 the generated output will not be the same.)
24367 __kernel void ramp(__write_only image2d_t dst,
24368 unsigned int index)
24370 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24373 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
24375 write_imagef(dst, loc, val);
24380 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
24382 __kernel void sierpinski_carpet(__write_only image2d_t dst,
24383 unsigned int index)
24385 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24387 float4 value = 0.0f;
24388 int x = loc.x + index;
24389 int y = loc.y + index;
24390 while (x > 0 || y > 0) {
24391 if (x % 3 == 1 && y % 3 == 1) {
24399 write_imagef(dst, loc, value);
24405 @section sierpinski
24407 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
24409 This source accepts the following options:
24413 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
24414 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
24417 Set frame rate, expressed as number of frames per second. Default
24421 Set seed which is used for random panning.
24424 Set max jump for single pan destination. Allowed range is from 1 to 10000.
24427 Set fractal type, can be default @code{carpet} or @code{triangle}.
24430 @c man end VIDEO SOURCES
24432 @chapter Video Sinks
24433 @c man begin VIDEO SINKS
24435 Below is a description of the currently available video sinks.
24437 @section buffersink
24439 Buffer video frames, and make them available to the end of the filter
24442 This sink is mainly intended for programmatic use, in particular
24443 through the interface defined in @file{libavfilter/buffersink.h}
24444 or the options system.
24446 It accepts a pointer to an AVBufferSinkContext structure, which
24447 defines the incoming buffers' formats, to be passed as the opaque
24448 parameter to @code{avfilter_init_filter} for initialization.
24452 Null video sink: do absolutely nothing with the input video. It is
24453 mainly useful as a template and for use in analysis / debugging
24456 @c man end VIDEO SINKS
24458 @chapter Multimedia Filters
24459 @c man begin MULTIMEDIA FILTERS
24461 Below is a description of the currently available multimedia filters.
24465 Convert input audio to a video output, displaying the audio bit scope.
24467 The filter accepts the following options:
24471 Set frame rate, expressed as number of frames per second. Default
24475 Specify the video size for the output. For the syntax of this option, check the
24476 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24477 Default value is @code{1024x256}.
24480 Specify list of colors separated by space or by '|' which will be used to
24481 draw channels. Unrecognized or missing colors will be replaced
24485 @section adrawgraph
24486 Draw a graph using input audio metadata.
24488 See @ref{drawgraph}
24490 @section agraphmonitor
24492 See @ref{graphmonitor}.
24494 @section ahistogram
24496 Convert input audio to a video output, displaying the volume histogram.
24498 The filter accepts the following options:
24502 Specify how histogram is calculated.
24504 It accepts the following values:
24507 Use single histogram for all channels.
24509 Use separate histogram for each channel.
24511 Default is @code{single}.
24514 Set frame rate, expressed as number of frames per second. Default
24518 Specify the video size for the output. For the syntax of this option, check the
24519 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24520 Default value is @code{hd720}.
24525 It accepts the following values:
24536 reverse logarithmic
24538 Default is @code{log}.
24541 Set amplitude scale.
24543 It accepts the following values:
24550 Default is @code{log}.
24553 Set how much frames to accumulate in histogram.
24554 Default is 1. Setting this to -1 accumulates all frames.
24557 Set histogram ratio of window height.
24560 Set sonogram sliding.
24562 It accepts the following values:
24565 replace old rows with new ones.
24567 scroll from top to bottom.
24569 Default is @code{replace}.
24572 @section aphasemeter
24574 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
24575 representing mean phase of current audio frame. A video output can also be produced and is
24576 enabled by default. The audio is passed through as first output.
24578 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
24579 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
24580 and @code{1} means channels are in phase.
24582 The filter accepts the following options, all related to its video output:
24586 Set the output frame rate. Default value is @code{25}.
24589 Set the video size for the output. For the syntax of this option, check the
24590 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24591 Default value is @code{800x400}.
24596 Specify the red, green, blue contrast. Default values are @code{2},
24597 @code{7} and @code{1}.
24598 Allowed range is @code{[0, 255]}.
24601 Set color which will be used for drawing median phase. If color is
24602 @code{none} which is default, no median phase value will be drawn.
24605 Enable video output. Default is enabled.
24608 @subsection phasing detection
24610 The filter also detects out of phase and mono sequences in stereo streams.
24611 It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
24613 The filter accepts the following options for this detection:
24617 Enable mono and out of phase detection. Default is disabled.
24620 Set phase tolerance for mono detection, in amplitude ratio. Default is @code{0}.
24621 Allowed range is @code{[0, 1]}.
24624 Set angle threshold for out of phase detection, in degree. Default is @code{170}.
24625 Allowed range is @code{[90, 180]}.
24628 Set mono or out of phase duration until notification, expressed in seconds. Default is @code{2}.
24631 @subsection Examples
24635 Complete example with @command{ffmpeg} to detect 1 second of mono with 0.001 phase tolerance:
24637 ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
24641 @section avectorscope
24643 Convert input audio to a video output, representing the audio vector
24646 The filter is used to measure the difference between channels of stereo
24647 audio stream. A monaural signal, consisting of identical left and right
24648 signal, results in straight vertical line. Any stereo separation is visible
24649 as a deviation from this line, creating a Lissajous figure.
24650 If the straight (or deviation from it) but horizontal line appears this
24651 indicates that the left and right channels are out of phase.
24653 The filter accepts the following options:
24657 Set the vectorscope mode.
24659 Available values are:
24662 Lissajous rotated by 45 degrees.
24665 Same as above but not rotated.
24668 Shape resembling half of circle.
24671 Default value is @samp{lissajous}.
24674 Set the video size for the output. For the syntax of this option, check the
24675 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24676 Default value is @code{400x400}.
24679 Set the output frame rate. Default value is @code{25}.
24685 Specify the red, green, blue and alpha contrast. Default values are @code{40},
24686 @code{160}, @code{80} and @code{255}.
24687 Allowed range is @code{[0, 255]}.
24693 Specify the red, green, blue and alpha fade. Default values are @code{15},
24694 @code{10}, @code{5} and @code{5}.
24695 Allowed range is @code{[0, 255]}.
24698 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
24699 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
24702 Set the vectorscope drawing mode.
24704 Available values are:
24707 Draw dot for each sample.
24710 Draw line between previous and current sample.
24713 Default value is @samp{dot}.
24716 Specify amplitude scale of audio samples.
24718 Available values are:
24734 Swap left channel axis with right channel axis.
24744 Mirror only x axis.
24747 Mirror only y axis.
24755 @subsection Examples
24759 Complete example using @command{ffplay}:
24761 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
24762 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
24766 @section bench, abench
24768 Benchmark part of a filtergraph.
24770 The filter accepts the following options:
24774 Start or stop a timer.
24776 Available values are:
24779 Get the current time, set it as frame metadata (using the key
24780 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
24783 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
24784 the input frame metadata to get the time difference. Time difference, average,
24785 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
24786 @code{min}) are then printed. The timestamps are expressed in seconds.
24790 @subsection Examples
24794 Benchmark @ref{selectivecolor} filter:
24796 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
24802 Concatenate audio and video streams, joining them together one after the
24805 The filter works on segments of synchronized video and audio streams. All
24806 segments must have the same number of streams of each type, and that will
24807 also be the number of streams at output.
24809 The filter accepts the following options:
24814 Set the number of segments. Default is 2.
24817 Set the number of output video streams, that is also the number of video
24818 streams in each segment. Default is 1.
24821 Set the number of output audio streams, that is also the number of audio
24822 streams in each segment. Default is 0.
24825 Activate unsafe mode: do not fail if segments have a different format.
24829 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
24830 @var{a} audio outputs.
24832 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
24833 segment, in the same order as the outputs, then the inputs for the second
24836 Related streams do not always have exactly the same duration, for various
24837 reasons including codec frame size or sloppy authoring. For that reason,
24838 related synchronized streams (e.g. a video and its audio track) should be
24839 concatenated at once. The concat filter will use the duration of the longest
24840 stream in each segment (except the last one), and if necessary pad shorter
24841 audio streams with silence.
24843 For this filter to work correctly, all segments must start at timestamp 0.
24845 All corresponding streams must have the same parameters in all segments; the
24846 filtering system will automatically select a common pixel format for video
24847 streams, and a common sample format, sample rate and channel layout for
24848 audio streams, but other settings, such as resolution, must be converted
24849 explicitly by the user.
24851 Different frame rates are acceptable but will result in variable frame rate
24852 at output; be sure to configure the output file to handle it.
24854 @subsection Examples
24858 Concatenate an opening, an episode and an ending, all in bilingual version
24859 (video in stream 0, audio in streams 1 and 2):
24861 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
24862 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
24863 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
24864 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
24868 Concatenate two parts, handling audio and video separately, using the
24869 (a)movie sources, and adjusting the resolution:
24871 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
24872 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
24873 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
24875 Note that a desync will happen at the stitch if the audio and video streams
24876 do not have exactly the same duration in the first file.
24880 @subsection Commands
24882 This filter supports the following commands:
24885 Close the current segment and step to the next one
24891 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
24892 level. By default, it logs a message at a frequency of 10Hz with the
24893 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
24894 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
24896 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
24897 sample format is double-precision floating point. The input stream will be converted to
24898 this specification, if needed. Users may need to insert aformat and/or aresample filters
24899 after this filter to obtain the original parameters.
24901 The filter also has a video output (see the @var{video} option) with a real
24902 time graph to observe the loudness evolution. The graphic contains the logged
24903 message mentioned above, so it is not printed anymore when this option is set,
24904 unless the verbose logging is set. The main graphing area contains the
24905 short-term loudness (3 seconds of analysis), and the gauge on the right is for
24906 the momentary loudness (400 milliseconds), but can optionally be configured
24907 to instead display short-term loudness (see @var{gauge}).
24909 The green area marks a +/- 1LU target range around the target loudness
24910 (-23LUFS by default, unless modified through @var{target}).
24912 More information about the Loudness Recommendation EBU R128 on
24913 @url{http://tech.ebu.ch/loudness}.
24915 The filter accepts the following options:
24920 Activate the video output. The audio stream is passed unchanged whether this
24921 option is set or no. The video stream will be the first output stream if
24922 activated. Default is @code{0}.
24925 Set the video size. This option is for video only. For the syntax of this
24927 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24928 Default and minimum resolution is @code{640x480}.
24931 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
24932 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
24933 other integer value between this range is allowed.
24936 Set metadata injection. If set to @code{1}, the audio input will be segmented
24937 into 100ms output frames, each of them containing various loudness information
24938 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
24940 Default is @code{0}.
24943 Force the frame logging level.
24945 Available values are:
24948 information logging level
24950 verbose logging level
24953 By default, the logging level is set to @var{info}. If the @option{video} or
24954 the @option{metadata} options are set, it switches to @var{verbose}.
24959 Available modes can be cumulated (the option is a @code{flag} type). Possible
24963 Disable any peak mode (default).
24965 Enable sample-peak mode.
24967 Simple peak mode looking for the higher sample value. It logs a message
24968 for sample-peak (identified by @code{SPK}).
24970 Enable true-peak mode.
24972 If enabled, the peak lookup is done on an over-sampled version of the input
24973 stream for better peak accuracy. It logs a message for true-peak.
24974 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
24975 This mode requires a build with @code{libswresample}.
24979 Treat mono input files as "dual mono". If a mono file is intended for playback
24980 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
24981 If set to @code{true}, this option will compensate for this effect.
24982 Multi-channel input files are not affected by this option.
24985 Set a specific pan law to be used for the measurement of dual mono files.
24986 This parameter is optional, and has a default value of -3.01dB.
24989 Set a specific target level (in LUFS) used as relative zero in the visualization.
24990 This parameter is optional and has a default value of -23LUFS as specified
24991 by EBU R128. However, material published online may prefer a level of -16LUFS
24992 (e.g. for use with podcasts or video platforms).
24995 Set the value displayed by the gauge. Valid values are @code{momentary} and s
24996 @code{shortterm}. By default the momentary value will be used, but in certain
24997 scenarios it may be more useful to observe the short term value instead (e.g.
25001 Sets the display scale for the loudness. Valid parameters are @code{absolute}
25002 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
25003 video output, not the summary or continuous log output.
25006 @subsection Examples
25010 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
25012 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
25016 Run an analysis with @command{ffmpeg}:
25018 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
25022 @section interleave, ainterleave
25024 Temporally interleave frames from several inputs.
25026 @code{interleave} works with video inputs, @code{ainterleave} with audio.
25028 These filters read frames from several inputs and send the oldest
25029 queued frame to the output.
25031 Input streams must have well defined, monotonically increasing frame
25034 In order to submit one frame to output, these filters need to enqueue
25035 at least one frame for each input, so they cannot work in case one
25036 input is not yet terminated and will not receive incoming frames.
25038 For example consider the case when one input is a @code{select} filter
25039 which always drops input frames. The @code{interleave} filter will keep
25040 reading from that input, but it will never be able to send new frames
25041 to output until the input sends an end-of-stream signal.
25043 Also, depending on inputs synchronization, the filters will drop
25044 frames in case one input receives more frames than the other ones, and
25045 the queue is already filled.
25047 These filters accept the following options:
25051 Set the number of different inputs, it is 2 by default.
25054 How to determine the end-of-stream.
25058 The duration of the longest input. (default)
25061 The duration of the shortest input.
25064 The duration of the first input.
25069 @subsection Examples
25073 Interleave frames belonging to different streams using @command{ffmpeg}:
25075 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
25079 Add flickering blur effect:
25081 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
25085 @section metadata, ametadata
25087 Manipulate frame metadata.
25089 This filter accepts the following options:
25093 Set mode of operation of the filter.
25095 Can be one of the following:
25099 If both @code{value} and @code{key} is set, select frames
25100 which have such metadata. If only @code{key} is set, select
25101 every frame that has such key in metadata.
25104 Add new metadata @code{key} and @code{value}. If key is already available
25108 Modify value of already present key.
25111 If @code{value} is set, delete only keys that have such value.
25112 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
25116 Print key and its value if metadata was found. If @code{key} is not set print all
25117 metadata values available in frame.
25121 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
25124 Set metadata value which will be used. This option is mandatory for
25125 @code{modify} and @code{add} mode.
25128 Which function to use when comparing metadata value and @code{value}.
25130 Can be one of following:
25134 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
25137 Values are interpreted as strings, returns true if metadata value starts with
25138 the @code{value} option string.
25141 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
25144 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
25147 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
25150 Values are interpreted as floats, returns true if expression from option @code{expr}
25154 Values are interpreted as strings, returns true if metadata value ends with
25155 the @code{value} option string.
25159 Set expression which is used when @code{function} is set to @code{expr}.
25160 The expression is evaluated through the eval API and can contain the following
25165 Float representation of @code{value} from metadata key.
25168 Float representation of @code{value} as supplied by user in @code{value} option.
25172 If specified in @code{print} mode, output is written to the named file. Instead of
25173 plain filename any writable url can be specified. Filename ``-'' is a shorthand
25174 for standard output. If @code{file} option is not set, output is written to the log
25175 with AV_LOG_INFO loglevel.
25178 Reduces buffering in print mode when output is written to a URL set using @var{file}.
25182 @subsection Examples
25186 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
25189 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
25192 Print silencedetect output to file @file{metadata.txt}.
25194 silencedetect,ametadata=mode=print:file=metadata.txt
25197 Direct all metadata to a pipe with file descriptor 4.
25199 metadata=mode=print:file='pipe\:4'
25203 @section perms, aperms
25205 Set read/write permissions for the output frames.
25207 These filters are mainly aimed at developers to test direct path in the
25208 following filter in the filtergraph.
25210 The filters accept the following options:
25214 Select the permissions mode.
25216 It accepts the following values:
25219 Do nothing. This is the default.
25221 Set all the output frames read-only.
25223 Set all the output frames directly writable.
25225 Make the frame read-only if writable, and writable if read-only.
25227 Set each output frame read-only or writable randomly.
25231 Set the seed for the @var{random} mode, must be an integer included between
25232 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
25233 @code{-1}, the filter will try to use a good random seed on a best effort
25237 Note: in case of auto-inserted filter between the permission filter and the
25238 following one, the permission might not be received as expected in that
25239 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
25240 perms/aperms filter can avoid this problem.
25242 @section realtime, arealtime
25244 Slow down filtering to match real time approximately.
25246 These filters will pause the filtering for a variable amount of time to
25247 match the output rate with the input timestamps.
25248 They are similar to the @option{re} option to @code{ffmpeg}.
25250 They accept the following options:
25254 Time limit for the pauses. Any pause longer than that will be considered
25255 a timestamp discontinuity and reset the timer. Default is 2 seconds.
25257 Speed factor for processing. The value must be a float larger than zero.
25258 Values larger than 1.0 will result in faster than realtime processing,
25259 smaller will slow processing down. The @var{limit} is automatically adapted
25260 accordingly. Default is 1.0.
25262 A processing speed faster than what is possible without these filters cannot
25267 @section select, aselect
25269 Select frames to pass in output.
25271 This filter accepts the following options:
25276 Set expression, which is evaluated for each input frame.
25278 If the expression is evaluated to zero, the frame is discarded.
25280 If the evaluation result is negative or NaN, the frame is sent to the
25281 first output; otherwise it is sent to the output with index
25282 @code{ceil(val)-1}, assuming that the input index starts from 0.
25284 For example a value of @code{1.2} corresponds to the output with index
25285 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
25288 Set the number of outputs. The output to which to send the selected
25289 frame is based on the result of the evaluation. Default value is 1.
25292 The expression can contain the following constants:
25296 The (sequential) number of the filtered frame, starting from 0.
25299 The (sequential) number of the selected frame, starting from 0.
25301 @item prev_selected_n
25302 The sequential number of the last selected frame. It's NAN if undefined.
25305 The timebase of the input timestamps.
25308 The PTS (Presentation TimeStamp) of the filtered video frame,
25309 expressed in @var{TB} units. It's NAN if undefined.
25312 The PTS of the filtered video frame,
25313 expressed in seconds. It's NAN if undefined.
25316 The PTS of the previously filtered video frame. It's NAN if undefined.
25318 @item prev_selected_pts
25319 The PTS of the last previously filtered video frame. It's NAN if undefined.
25321 @item prev_selected_t
25322 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
25325 The PTS of the first video frame in the video. It's NAN if undefined.
25328 The time of the first video frame in the video. It's NAN if undefined.
25330 @item pict_type @emph{(video only)}
25331 The type of the filtered frame. It can assume one of the following
25343 @item interlace_type @emph{(video only)}
25344 The frame interlace type. It can assume one of the following values:
25347 The frame is progressive (not interlaced).
25349 The frame is top-field-first.
25351 The frame is bottom-field-first.
25354 @item consumed_sample_n @emph{(audio only)}
25355 the number of selected samples before the current frame
25357 @item samples_n @emph{(audio only)}
25358 the number of samples in the current frame
25360 @item sample_rate @emph{(audio only)}
25361 the input sample rate
25364 This is 1 if the filtered frame is a key-frame, 0 otherwise.
25367 the position in the file of the filtered frame, -1 if the information
25368 is not available (e.g. for synthetic video)
25370 @item scene @emph{(video only)}
25371 value between 0 and 1 to indicate a new scene; a low value reflects a low
25372 probability for the current frame to introduce a new scene, while a higher
25373 value means the current frame is more likely to be one (see the example below)
25375 @item concatdec_select
25376 The concat demuxer can select only part of a concat input file by setting an
25377 inpoint and an outpoint, but the output packets may not be entirely contained
25378 in the selected interval. By using this variable, it is possible to skip frames
25379 generated by the concat demuxer which are not exactly contained in the selected
25382 This works by comparing the frame pts against the @var{lavf.concat.start_time}
25383 and the @var{lavf.concat.duration} packet metadata values which are also
25384 present in the decoded frames.
25386 The @var{concatdec_select} variable is -1 if the frame pts is at least
25387 start_time and either the duration metadata is missing or the frame pts is less
25388 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
25391 That basically means that an input frame is selected if its pts is within the
25392 interval set by the concat demuxer.
25396 The default value of the select expression is "1".
25398 @subsection Examples
25402 Select all frames in input:
25407 The example above is the same as:
25419 Select only I-frames:
25421 select='eq(pict_type\,I)'
25425 Select one frame every 100:
25427 select='not(mod(n\,100))'
25431 Select only frames contained in the 10-20 time interval:
25433 select=between(t\,10\,20)
25437 Select only I-frames contained in the 10-20 time interval:
25439 select=between(t\,10\,20)*eq(pict_type\,I)
25443 Select frames with a minimum distance of 10 seconds:
25445 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
25449 Use aselect to select only audio frames with samples number > 100:
25451 aselect='gt(samples_n\,100)'
25455 Create a mosaic of the first scenes:
25457 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
25460 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
25464 Send even and odd frames to separate outputs, and compose them:
25466 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
25470 Select useful frames from an ffconcat file which is using inpoints and
25471 outpoints but where the source files are not intra frame only.
25473 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
25477 @section sendcmd, asendcmd
25479 Send commands to filters in the filtergraph.
25481 These filters read commands to be sent to other filters in the
25484 @code{sendcmd} must be inserted between two video filters,
25485 @code{asendcmd} must be inserted between two audio filters, but apart
25486 from that they act the same way.
25488 The specification of commands can be provided in the filter arguments
25489 with the @var{commands} option, or in a file specified by the
25490 @var{filename} option.
25492 These filters accept the following options:
25495 Set the commands to be read and sent to the other filters.
25497 Set the filename of the commands to be read and sent to the other
25501 @subsection Commands syntax
25503 A commands description consists of a sequence of interval
25504 specifications, comprising a list of commands to be executed when a
25505 particular event related to that interval occurs. The occurring event
25506 is typically the current frame time entering or leaving a given time
25509 An interval is specified by the following syntax:
25511 @var{START}[-@var{END}] @var{COMMANDS};
25514 The time interval is specified by the @var{START} and @var{END} times.
25515 @var{END} is optional and defaults to the maximum time.
25517 The current frame time is considered within the specified interval if
25518 it is included in the interval [@var{START}, @var{END}), that is when
25519 the time is greater or equal to @var{START} and is lesser than
25522 @var{COMMANDS} consists of a sequence of one or more command
25523 specifications, separated by ",", relating to that interval. The
25524 syntax of a command specification is given by:
25526 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
25529 @var{FLAGS} is optional and specifies the type of events relating to
25530 the time interval which enable sending the specified command, and must
25531 be a non-null sequence of identifier flags separated by "+" or "|" and
25532 enclosed between "[" and "]".
25534 The following flags are recognized:
25537 The command is sent when the current frame timestamp enters the
25538 specified interval. In other words, the command is sent when the
25539 previous frame timestamp was not in the given interval, and the
25543 The command is sent when the current frame timestamp leaves the
25544 specified interval. In other words, the command is sent when the
25545 previous frame timestamp was in the given interval, and the
25549 The command @var{ARG} is interpreted as expression and result of
25550 expression is passed as @var{ARG}.
25552 The expression is evaluated through the eval API and can contain the following
25557 Original position in the file of the frame, or undefined if undefined
25558 for the current frame.
25561 The presentation timestamp in input.
25564 The count of the input frame for video or audio, starting from 0.
25567 The time in seconds of the current frame.
25570 The start time in seconds of the current command interval.
25573 The end time in seconds of the current command interval.
25576 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
25581 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
25584 @var{TARGET} specifies the target of the command, usually the name of
25585 the filter class or a specific filter instance name.
25587 @var{COMMAND} specifies the name of the command for the target filter.
25589 @var{ARG} is optional and specifies the optional list of argument for
25590 the given @var{COMMAND}.
25592 Between one interval specification and another, whitespaces, or
25593 sequences of characters starting with @code{#} until the end of line,
25594 are ignored and can be used to annotate comments.
25596 A simplified BNF description of the commands specification syntax
25599 @var{COMMAND_FLAG} ::= "enter" | "leave"
25600 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
25601 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
25602 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
25603 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
25604 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
25607 @subsection Examples
25611 Specify audio tempo change at second 4:
25613 asendcmd=c='4.0 atempo tempo 1.5',atempo
25617 Target a specific filter instance:
25619 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
25623 Specify a list of drawtext and hue commands in a file.
25625 # show text in the interval 5-10
25626 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
25627 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
25629 # desaturate the image in the interval 15-20
25630 15.0-20.0 [enter] hue s 0,
25631 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
25633 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
25635 # apply an exponential saturation fade-out effect, starting from time 25
25636 25 [enter] hue s exp(25-t)
25639 A filtergraph allowing to read and process the above command list
25640 stored in a file @file{test.cmd}, can be specified with:
25642 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
25647 @section setpts, asetpts
25649 Change the PTS (presentation timestamp) of the input frames.
25651 @code{setpts} works on video frames, @code{asetpts} on audio frames.
25653 This filter accepts the following options:
25658 The expression which is evaluated for each frame to construct its timestamp.
25662 The expression is evaluated through the eval API and can contain the following
25666 @item FRAME_RATE, FR
25667 frame rate, only defined for constant frame-rate video
25670 The presentation timestamp in input
25673 The count of the input frame for video or the number of consumed samples,
25674 not including the current frame for audio, starting from 0.
25676 @item NB_CONSUMED_SAMPLES
25677 The number of consumed samples, not including the current frame (only
25680 @item NB_SAMPLES, S
25681 The number of samples in the current frame (only audio)
25683 @item SAMPLE_RATE, SR
25684 The audio sample rate.
25687 The PTS of the first frame.
25690 the time in seconds of the first frame
25693 State whether the current frame is interlaced.
25696 the time in seconds of the current frame
25699 original position in the file of the frame, or undefined if undefined
25700 for the current frame
25703 The previous input PTS.
25706 previous input time in seconds
25709 The previous output PTS.
25712 previous output time in seconds
25715 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
25719 The wallclock (RTC) time at the start of the movie in microseconds.
25722 The timebase of the input timestamps.
25726 @subsection Examples
25730 Start counting PTS from zero
25732 setpts=PTS-STARTPTS
25736 Apply fast motion effect:
25742 Apply slow motion effect:
25748 Set fixed rate of 25 frames per second:
25754 Set fixed rate 25 fps with some jitter:
25756 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
25760 Apply an offset of 10 seconds to the input PTS:
25766 Generate timestamps from a "live source" and rebase onto the current timebase:
25768 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
25772 Generate timestamps by counting samples:
25781 Force color range for the output video frame.
25783 The @code{setrange} filter marks the color range property for the
25784 output frames. It does not change the input frame, but only sets the
25785 corresponding property, which affects how the frame is treated by
25788 The filter accepts the following options:
25793 Available values are:
25797 Keep the same color range property.
25799 @item unspecified, unknown
25800 Set the color range as unspecified.
25802 @item limited, tv, mpeg
25803 Set the color range as limited.
25805 @item full, pc, jpeg
25806 Set the color range as full.
25810 @section settb, asettb
25812 Set the timebase to use for the output frames timestamps.
25813 It is mainly useful for testing timebase configuration.
25815 It accepts the following parameters:
25820 The expression which is evaluated into the output timebase.
25824 The value for @option{tb} is an arithmetic expression representing a
25825 rational. The expression can contain the constants "AVTB" (the default
25826 timebase), "intb" (the input timebase) and "sr" (the sample rate,
25827 audio only). Default value is "intb".
25829 @subsection Examples
25833 Set the timebase to 1/25:
25839 Set the timebase to 1/10:
25845 Set the timebase to 1001/1000:
25851 Set the timebase to 2*intb:
25857 Set the default timebase value:
25864 Convert input audio to a video output representing frequency spectrum
25865 logarithmically using Brown-Puckette constant Q transform algorithm with
25866 direct frequency domain coefficient calculation (but the transform itself
25867 is not really constant Q, instead the Q factor is actually variable/clamped),
25868 with musical tone scale, from E0 to D#10.
25870 The filter accepts the following options:
25874 Specify the video size for the output. It must be even. For the syntax of this option,
25875 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25876 Default value is @code{1920x1080}.
25879 Set the output frame rate. Default value is @code{25}.
25882 Set the bargraph height. It must be even. Default value is @code{-1} which
25883 computes the bargraph height automatically.
25886 Set the axis height. It must be even. Default value is @code{-1} which computes
25887 the axis height automatically.
25890 Set the sonogram height. It must be even. Default value is @code{-1} which
25891 computes the sonogram height automatically.
25894 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
25895 instead. Default value is @code{1}.
25897 @item sono_v, volume
25898 Specify the sonogram volume expression. It can contain variables:
25901 the @var{bar_v} evaluated expression
25902 @item frequency, freq, f
25903 the frequency where it is evaluated
25904 @item timeclamp, tc
25905 the value of @var{timeclamp} option
25909 @item a_weighting(f)
25910 A-weighting of equal loudness
25911 @item b_weighting(f)
25912 B-weighting of equal loudness
25913 @item c_weighting(f)
25914 C-weighting of equal loudness.
25916 Default value is @code{16}.
25918 @item bar_v, volume2
25919 Specify the bargraph volume expression. It can contain variables:
25922 the @var{sono_v} evaluated expression
25923 @item frequency, freq, f
25924 the frequency where it is evaluated
25925 @item timeclamp, tc
25926 the value of @var{timeclamp} option
25930 @item a_weighting(f)
25931 A-weighting of equal loudness
25932 @item b_weighting(f)
25933 B-weighting of equal loudness
25934 @item c_weighting(f)
25935 C-weighting of equal loudness.
25937 Default value is @code{sono_v}.
25939 @item sono_g, gamma
25940 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
25941 higher gamma makes the spectrum having more range. Default value is @code{3}.
25942 Acceptable range is @code{[1, 7]}.
25944 @item bar_g, gamma2
25945 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
25949 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
25950 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
25952 @item timeclamp, tc
25953 Specify the transform timeclamp. At low frequency, there is trade-off between
25954 accuracy in time domain and frequency domain. If timeclamp is lower,
25955 event in time domain is represented more accurately (such as fast bass drum),
25956 otherwise event in frequency domain is represented more accurately
25957 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
25960 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
25961 limits future samples by applying asymmetric windowing in time domain, useful
25962 when low latency is required. Accepted range is @code{[0, 1]}.
25965 Specify the transform base frequency. Default value is @code{20.01523126408007475},
25966 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
25969 Specify the transform end frequency. Default value is @code{20495.59681441799654},
25970 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
25973 This option is deprecated and ignored.
25976 Specify the transform length in time domain. Use this option to control accuracy
25977 trade-off between time domain and frequency domain at every frequency sample.
25978 It can contain variables:
25980 @item frequency, freq, f
25981 the frequency where it is evaluated
25982 @item timeclamp, tc
25983 the value of @var{timeclamp} option.
25985 Default value is @code{384*tc/(384+tc*f)}.
25988 Specify the transform count for every video frame. Default value is @code{6}.
25989 Acceptable range is @code{[1, 30]}.
25992 Specify the transform count for every single pixel. Default value is @code{0},
25993 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
25996 Specify font file for use with freetype to draw the axis. If not specified,
25997 use embedded font. Note that drawing with font file or embedded font is not
25998 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
26002 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
26003 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
26007 Specify font color expression. This is arithmetic expression that should return
26008 integer value 0xRRGGBB. It can contain variables:
26010 @item frequency, freq, f
26011 the frequency where it is evaluated
26012 @item timeclamp, tc
26013 the value of @var{timeclamp} option
26018 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
26019 @item r(x), g(x), b(x)
26020 red, green, and blue value of intensity x.
26022 Default value is @code{st(0, (midi(f)-59.5)/12);
26023 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
26024 r(1-ld(1)) + b(ld(1))}.
26027 Specify image file to draw the axis. This option override @var{fontfile} and
26028 @var{fontcolor} option.
26031 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
26032 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
26033 Default value is @code{1}.
26036 Set colorspace. The accepted values are:
26039 Unspecified (default)
26048 BT.470BG or BT.601-6 625
26051 SMPTE-170M or BT.601-6 525
26057 BT.2020 with non-constant luminance
26062 Set spectrogram color scheme. This is list of floating point values with format
26063 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
26064 The default is @code{1|0.5|0|0|0.5|1}.
26068 @subsection Examples
26072 Playing audio while showing the spectrum:
26074 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
26078 Same as above, but with frame rate 30 fps:
26080 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
26084 Playing at 1280x720:
26086 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
26090 Disable sonogram display:
26096 A1 and its harmonics: A1, A2, (near)E3, A3:
26098 ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
26099 asplit[a][out1]; [a] showcqt [out0]'
26103 Same as above, but with more accuracy in frequency domain:
26105 ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
26106 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
26112 bar_v=10:sono_v=bar_v*a_weighting(f)
26116 Custom gamma, now spectrum is linear to the amplitude.
26122 Custom tlength equation:
26124 tc=0.33:tlength='st(0,0.17); 384*tc / (384 / ld(0) + tc*f /(1-ld(0))) + 384*tc / (tc*f / ld(0) + 384 /(1-ld(0)))'
26128 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
26130 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
26134 Custom font using fontconfig:
26136 font='Courier New,Monospace,mono|bold'
26140 Custom frequency range with custom axis using image file:
26142 axisfile=myaxis.png:basefreq=40:endfreq=10000
26148 Convert input audio to video output representing the audio power spectrum.
26149 Audio amplitude is on Y-axis while frequency is on X-axis.
26151 The filter accepts the following options:
26155 Specify size of video. For the syntax of this option, check the
26156 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26157 Default is @code{1024x512}.
26161 This set how each frequency bin will be represented.
26163 It accepts the following values:
26169 Default is @code{bar}.
26172 Set amplitude scale.
26174 It accepts the following values:
26188 Default is @code{log}.
26191 Set frequency scale.
26193 It accepts the following values:
26202 Reverse logarithmic scale.
26204 Default is @code{lin}.
26207 Set window size. Allowed range is from 16 to 65536.
26209 Default is @code{2048}
26212 Set windowing function.
26214 It accepts the following values:
26237 Default is @code{hanning}.
26240 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26241 which means optimal overlap for selected window function will be picked.
26244 Set time averaging. Setting this to 0 will display current maximal peaks.
26245 Default is @code{1}, which means time averaging is disabled.
26248 Specify list of colors separated by space or by '|' which will be used to
26249 draw channel frequencies. Unrecognized or missing colors will be replaced
26253 Set channel display mode.
26255 It accepts the following values:
26260 Default is @code{combined}.
26263 Set minimum amplitude used in @code{log} amplitude scaler.
26266 Set data display mode.
26268 It accepts the following values:
26274 Default is @code{magnitude}.
26277 @section showspatial
26279 Convert stereo input audio to a video output, representing the spatial relationship
26280 between two channels.
26282 The filter accepts the following options:
26286 Specify the video size for the output. For the syntax of this option, check the
26287 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26288 Default value is @code{512x512}.
26291 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
26294 Set window function.
26296 It accepts the following values:
26321 Default value is @code{hann}.
26324 Set ratio of overlap window. Default value is @code{0.5}.
26325 When value is @code{1} overlap is set to recommended size for specific
26326 window function currently used.
26329 @anchor{showspectrum}
26330 @section showspectrum
26332 Convert input audio to a video output, representing the audio frequency
26335 The filter accepts the following options:
26339 Specify the video size for the output. For the syntax of this option, check the
26340 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26341 Default value is @code{640x512}.
26344 Specify how the spectrum should slide along the window.
26346 It accepts the following values:
26349 the samples start again on the left when they reach the right
26351 the samples scroll from right to left
26353 frames are only produced when the samples reach the right
26355 the samples scroll from left to right
26358 Default value is @code{replace}.
26361 Specify display mode.
26363 It accepts the following values:
26366 all channels are displayed in the same row
26368 all channels are displayed in separate rows
26371 Default value is @samp{combined}.
26374 Specify display color mode.
26376 It accepts the following values:
26379 each channel is displayed in a separate color
26381 each channel is displayed using the same color scheme
26383 each channel is displayed using the rainbow color scheme
26385 each channel is displayed using the moreland color scheme
26387 each channel is displayed using the nebulae color scheme
26389 each channel is displayed using the fire color scheme
26391 each channel is displayed using the fiery color scheme
26393 each channel is displayed using the fruit color scheme
26395 each channel is displayed using the cool color scheme
26397 each channel is displayed using the magma color scheme
26399 each channel is displayed using the green color scheme
26401 each channel is displayed using the viridis color scheme
26403 each channel is displayed using the plasma color scheme
26405 each channel is displayed using the cividis color scheme
26407 each channel is displayed using the terrain color scheme
26410 Default value is @samp{channel}.
26413 Specify scale used for calculating intensity color values.
26415 It accepts the following values:
26420 square root, default
26431 Default value is @samp{sqrt}.
26434 Specify frequency scale.
26436 It accepts the following values:
26444 Default value is @samp{lin}.
26447 Set saturation modifier for displayed colors. Negative values provide
26448 alternative color scheme. @code{0} is no saturation at all.
26449 Saturation must be in [-10.0, 10.0] range.
26450 Default value is @code{1}.
26453 Set window function.
26455 It accepts the following values:
26480 Default value is @code{hann}.
26483 Set orientation of time vs frequency axis. Can be @code{vertical} or
26484 @code{horizontal}. Default is @code{vertical}.
26487 Set ratio of overlap window. Default value is @code{0}.
26488 When value is @code{1} overlap is set to recommended size for specific
26489 window function currently used.
26492 Set scale gain for calculating intensity color values.
26493 Default value is @code{1}.
26496 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
26499 Set color rotation, must be in [-1.0, 1.0] range.
26500 Default value is @code{0}.
26503 Set start frequency from which to display spectrogram. Default is @code{0}.
26506 Set stop frequency to which to display spectrogram. Default is @code{0}.
26509 Set upper frame rate limit. Default is @code{auto}, unlimited.
26512 Draw time and frequency axes and legends. Default is disabled.
26515 The usage is very similar to the showwaves filter; see the examples in that
26518 @subsection Examples
26522 Large window with logarithmic color scaling:
26524 showspectrum=s=1280x480:scale=log
26528 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
26530 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
26531 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
26535 @section showspectrumpic
26537 Convert input audio to a single video frame, representing the audio frequency
26540 The filter accepts the following options:
26544 Specify the video size for the output. For the syntax of this option, check the
26545 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26546 Default value is @code{4096x2048}.
26549 Specify display mode.
26551 It accepts the following values:
26554 all channels are displayed in the same row
26556 all channels are displayed in separate rows
26558 Default value is @samp{combined}.
26561 Specify display color mode.
26563 It accepts the following values:
26566 each channel is displayed in a separate color
26568 each channel is displayed using the same color scheme
26570 each channel is displayed using the rainbow color scheme
26572 each channel is displayed using the moreland color scheme
26574 each channel is displayed using the nebulae color scheme
26576 each channel is displayed using the fire color scheme
26578 each channel is displayed using the fiery color scheme
26580 each channel is displayed using the fruit color scheme
26582 each channel is displayed using the cool color scheme
26584 each channel is displayed using the magma color scheme
26586 each channel is displayed using the green color scheme
26588 each channel is displayed using the viridis color scheme
26590 each channel is displayed using the plasma color scheme
26592 each channel is displayed using the cividis color scheme
26594 each channel is displayed using the terrain color scheme
26596 Default value is @samp{intensity}.
26599 Specify scale used for calculating intensity color values.
26601 It accepts the following values:
26606 square root, default
26616 Default value is @samp{log}.
26619 Specify frequency scale.
26621 It accepts the following values:
26629 Default value is @samp{lin}.
26632 Set saturation modifier for displayed colors. Negative values provide
26633 alternative color scheme. @code{0} is no saturation at all.
26634 Saturation must be in [-10.0, 10.0] range.
26635 Default value is @code{1}.
26638 Set window function.
26640 It accepts the following values:
26664 Default value is @code{hann}.
26667 Set orientation of time vs frequency axis. Can be @code{vertical} or
26668 @code{horizontal}. Default is @code{vertical}.
26671 Set scale gain for calculating intensity color values.
26672 Default value is @code{1}.
26675 Draw time and frequency axes and legends. Default is enabled.
26678 Set color rotation, must be in [-1.0, 1.0] range.
26679 Default value is @code{0}.
26682 Set start frequency from which to display spectrogram. Default is @code{0}.
26685 Set stop frequency to which to display spectrogram. Default is @code{0}.
26688 @subsection Examples
26692 Extract an audio spectrogram of a whole audio track
26693 in a 1024x1024 picture using @command{ffmpeg}:
26695 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
26699 @section showvolume
26701 Convert input audio volume to a video output.
26703 The filter accepts the following options:
26710 Set border width, allowed range is [0, 5]. Default is 1.
26713 Set channel width, allowed range is [80, 8192]. Default is 400.
26716 Set channel height, allowed range is [1, 900]. Default is 20.
26719 Set fade, allowed range is [0, 1]. Default is 0.95.
26722 Set volume color expression.
26724 The expression can use the following variables:
26728 Current max volume of channel in dB.
26734 Current channel number, starting from 0.
26738 If set, displays channel names. Default is enabled.
26741 If set, displays volume values. Default is enabled.
26744 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
26745 default is @code{h}.
26748 Set step size, allowed range is [0, 5]. Default is 0, which means
26752 Set background opacity, allowed range is [0, 1]. Default is 0.
26755 Set metering mode, can be peak: @code{p} or rms: @code{r},
26756 default is @code{p}.
26759 Set display scale, can be linear: @code{lin} or log: @code{log},
26760 default is @code{lin}.
26764 If set to > 0., display a line for the max level
26765 in the previous seconds.
26766 default is disabled: @code{0.}
26769 The color of the max line. Use when @code{dm} option is set to > 0.
26770 default is: @code{orange}
26775 Convert input audio to a video output, representing the samples waves.
26777 The filter accepts the following options:
26781 Specify the video size for the output. For the syntax of this option, check the
26782 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26783 Default value is @code{600x240}.
26788 Available values are:
26791 Draw a point for each sample.
26794 Draw a vertical line for each sample.
26797 Draw a point for each sample and a line between them.
26800 Draw a centered vertical line for each sample.
26803 Default value is @code{point}.
26806 Set the number of samples which are printed on the same column. A
26807 larger value will decrease the frame rate. Must be a positive
26808 integer. This option can be set only if the value for @var{rate}
26809 is not explicitly specified.
26812 Set the (approximate) output frame rate. This is done by setting the
26813 option @var{n}. Default value is "25".
26815 @item split_channels
26816 Set if channels should be drawn separately or overlap. Default value is 0.
26819 Set colors separated by '|' which are going to be used for drawing of each channel.
26822 Set amplitude scale.
26824 Available values are:
26842 Set the draw mode. This is mostly useful to set for high @var{n}.
26844 Available values are:
26847 Scale pixel values for each drawn sample.
26850 Draw every sample directly.
26853 Default value is @code{scale}.
26856 @subsection Examples
26860 Output the input file audio and the corresponding video representation
26863 amovie=a.mp3,asplit[out0],showwaves[out1]
26867 Create a synthetic signal and show it with showwaves, forcing a
26868 frame rate of 30 frames per second:
26870 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
26874 @section showwavespic
26876 Convert input audio to a single video frame, representing the samples waves.
26878 The filter accepts the following options:
26882 Specify the video size for the output. For the syntax of this option, check the
26883 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26884 Default value is @code{600x240}.
26886 @item split_channels
26887 Set if channels should be drawn separately or overlap. Default value is 0.
26890 Set colors separated by '|' which are going to be used for drawing of each channel.
26893 Set amplitude scale.
26895 Available values are:
26915 Available values are:
26918 Scale pixel values for each drawn sample.
26921 Draw every sample directly.
26924 Default value is @code{scale}.
26927 Set the filter mode.
26929 Available values are:
26932 Use average samples values for each drawn sample.
26935 Use peak samples values for each drawn sample.
26938 Default value is @code{average}.
26941 @subsection Examples
26945 Extract a channel split representation of the wave form of a whole audio track
26946 in a 1024x800 picture using @command{ffmpeg}:
26948 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
26952 @section sidedata, asidedata
26954 Delete frame side data, or select frames based on it.
26956 This filter accepts the following options:
26960 Set mode of operation of the filter.
26962 Can be one of the following:
26966 Select every frame with side data of @code{type}.
26969 Delete side data of @code{type}. If @code{type} is not set, delete all side
26975 Set side data type used with all modes. Must be set for @code{select} mode. For
26976 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
26977 in @file{libavutil/frame.h}. For example, to choose
26978 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
26982 @section spectrumsynth
26984 Synthesize audio from 2 input video spectrums, first input stream represents
26985 magnitude across time and second represents phase across time.
26986 The filter will transform from frequency domain as displayed in videos back
26987 to time domain as presented in audio output.
26989 This filter is primarily created for reversing processed @ref{showspectrum}
26990 filter outputs, but can synthesize sound from other spectrograms too.
26991 But in such case results are going to be poor if the phase data is not
26992 available, because in such cases phase data need to be recreated, usually
26993 it's just recreated from random noise.
26994 For best results use gray only output (@code{channel} color mode in
26995 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
26996 @code{lin} scale for phase video. To produce phase, for 2nd video, use
26997 @code{data} option. Inputs videos should generally use @code{fullframe}
26998 slide mode as that saves resources needed for decoding video.
27000 The filter accepts the following options:
27004 Specify sample rate of output audio, the sample rate of audio from which
27005 spectrum was generated may differ.
27008 Set number of channels represented in input video spectrums.
27011 Set scale which was used when generating magnitude input spectrum.
27012 Can be @code{lin} or @code{log}. Default is @code{log}.
27015 Set slide which was used when generating inputs spectrums.
27016 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
27017 Default is @code{fullframe}.
27020 Set window function used for resynthesis.
27023 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
27024 which means optimal overlap for selected window function will be picked.
27027 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
27028 Default is @code{vertical}.
27031 @subsection Examples
27035 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
27036 then resynthesize videos back to audio with spectrumsynth:
27038 ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=log:overlap=0.875:color=channel:slide=fullframe:data=magnitude -an -c:v rawvideo magnitude.nut
27039 ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=lin:overlap=0.875:color=channel:slide=fullframe:data=phase -an -c:v rawvideo phase.nut
27040 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
27044 @section split, asplit
27046 Split input into several identical outputs.
27048 @code{asplit} works with audio input, @code{split} with video.
27050 The filter accepts a single parameter which specifies the number of outputs. If
27051 unspecified, it defaults to 2.
27053 @subsection Examples
27057 Create two separate outputs from the same input:
27059 [in] split [out0][out1]
27063 To create 3 or more outputs, you need to specify the number of
27066 [in] asplit=3 [out0][out1][out2]
27070 Create two separate outputs from the same input, one cropped and
27073 [in] split [splitout1][splitout2];
27074 [splitout1] crop=100:100:0:0 [cropout];
27075 [splitout2] pad=200:200:100:100 [padout];
27079 Create 5 copies of the input audio with @command{ffmpeg}:
27081 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
27087 Receive commands sent through a libzmq client, and forward them to
27088 filters in the filtergraph.
27090 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
27091 must be inserted between two video filters, @code{azmq} between two
27092 audio filters. Both are capable to send messages to any filter type.
27094 To enable these filters you need to install the libzmq library and
27095 headers and configure FFmpeg with @code{--enable-libzmq}.
27097 For more information about libzmq see:
27098 @url{http://www.zeromq.org/}
27100 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
27101 receives messages sent through a network interface defined by the
27102 @option{bind_address} (or the abbreviation "@option{b}") option.
27103 Default value of this option is @file{tcp://localhost:5555}. You may
27104 want to alter this value to your needs, but do not forget to escape any
27105 ':' signs (see @ref{filtergraph escaping}).
27107 The received message must be in the form:
27109 @var{TARGET} @var{COMMAND} [@var{ARG}]
27112 @var{TARGET} specifies the target of the command, usually the name of
27113 the filter class or a specific filter instance name. The default
27114 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
27115 but you can override this by using the @samp{filter_name@@id} syntax
27116 (see @ref{Filtergraph syntax}).
27118 @var{COMMAND} specifies the name of the command for the target filter.
27120 @var{ARG} is optional and specifies the optional argument list for the
27121 given @var{COMMAND}.
27123 Upon reception, the message is processed and the corresponding command
27124 is injected into the filtergraph. Depending on the result, the filter
27125 will send a reply to the client, adopting the format:
27127 @var{ERROR_CODE} @var{ERROR_REASON}
27131 @var{MESSAGE} is optional.
27133 @subsection Examples
27135 Look at @file{tools/zmqsend} for an example of a zmq client which can
27136 be used to send commands processed by these filters.
27138 Consider the following filtergraph generated by @command{ffplay}.
27139 In this example the last overlay filter has an instance name. All other
27140 filters will have default instance names.
27143 ffplay -dumpgraph 1 -f lavfi "
27144 color=s=100x100:c=red [l];
27145 color=s=100x100:c=blue [r];
27146 nullsrc=s=200x100, zmq [bg];
27147 [bg][l] overlay [bg+l];
27148 [bg+l][r] overlay@@my=x=100 "
27151 To change the color of the left side of the video, the following
27152 command can be used:
27154 echo Parsed_color_0 c yellow | tools/zmqsend
27157 To change the right side:
27159 echo Parsed_color_1 c pink | tools/zmqsend
27162 To change the position of the right side:
27164 echo overlay@@my x 150 | tools/zmqsend
27168 @c man end MULTIMEDIA FILTERS
27170 @chapter Multimedia Sources
27171 @c man begin MULTIMEDIA SOURCES
27173 Below is a description of the currently available multimedia sources.
27177 This is the same as @ref{movie} source, except it selects an audio
27183 Read audio and/or video stream(s) from a movie container.
27185 It accepts the following parameters:
27189 The name of the resource to read (not necessarily a file; it can also be a
27190 device or a stream accessed through some protocol).
27192 @item format_name, f
27193 Specifies the format assumed for the movie to read, and can be either
27194 the name of a container or an input device. If not specified, the
27195 format is guessed from @var{movie_name} or by probing.
27197 @item seek_point, sp
27198 Specifies the seek point in seconds. The frames will be output
27199 starting from this seek point. The parameter is evaluated with
27200 @code{av_strtod}, so the numerical value may be suffixed by an IS
27201 postfix. The default value is "0".
27204 Specifies the streams to read. Several streams can be specified,
27205 separated by "+". The source will then have as many outputs, in the
27206 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
27207 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
27208 respectively the default (best suited) video and audio stream. Default
27209 is "dv", or "da" if the filter is called as "amovie".
27211 @item stream_index, si
27212 Specifies the index of the video stream to read. If the value is -1,
27213 the most suitable video stream will be automatically selected. The default
27214 value is "-1". Deprecated. If the filter is called "amovie", it will select
27215 audio instead of video.
27218 Specifies how many times to read the stream in sequence.
27219 If the value is 0, the stream will be looped infinitely.
27220 Default value is "1".
27222 Note that when the movie is looped the source timestamps are not
27223 changed, so it will generate non monotonically increasing timestamps.
27225 @item discontinuity
27226 Specifies the time difference between frames above which the point is
27227 considered a timestamp discontinuity which is removed by adjusting the later
27231 It allows overlaying a second video on top of the main input of
27232 a filtergraph, as shown in this graph:
27234 input -----------> deltapts0 --> overlay --> output
27237 movie --> scale--> deltapts1 -------+
27239 @subsection Examples
27243 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
27244 on top of the input labelled "in":
27246 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
27247 [in] setpts=PTS-STARTPTS [main];
27248 [main][over] overlay=16:16 [out]
27252 Read from a video4linux2 device, and overlay it on top of the input
27255 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
27256 [in] setpts=PTS-STARTPTS [main];
27257 [main][over] overlay=16:16 [out]
27261 Read the first video stream and the audio stream with id 0x81 from
27262 dvd.vob; the video is connected to the pad named "video" and the audio is
27263 connected to the pad named "audio":
27265 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
27269 @subsection Commands
27271 Both movie and amovie support the following commands:
27274 Perform seek using "av_seek_frame".
27275 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
27278 @var{stream_index}: If stream_index is -1, a default
27279 stream is selected, and @var{timestamp} is automatically converted
27280 from AV_TIME_BASE units to the stream specific time_base.
27282 @var{timestamp}: Timestamp in AVStream.time_base units
27283 or, if no stream is specified, in AV_TIME_BASE units.
27285 @var{flags}: Flags which select direction and seeking mode.
27289 Get movie duration in AV_TIME_BASE units.
27293 @c man end MULTIMEDIA SOURCES